20201215.md

ArXiv cs.CV --Tue, 15 Dec 2020

1.Real-Time High-Resolution Background Matting ⬇️

We introduce a real-time, high-resolution background replacement technique which operates at 30fps in 4K resolution, and 60fps for HD on a modern GPU. Our technique is based on background matting, where an additional frame of the background is captured and used in recovering the alpha matte and the foreground layer. The main challenge is to compute a high-quality alpha matte, preserving strand-level hair details, while processing high-resolution images in real-time. To achieve this goal, we employ two neural networks; a base network computes a low-resolution result which is refined by a second network operating at high-resolution on selective patches. We introduce two largescale video and image matting datasets: VideoMatte240K and PhotoMatte13K/85. Our approach yields higher quality results compared to the previous state-of-the-art in background matting, while simultaneously yielding a dramatic boost in both speed and resolution.

2.img2pose: Face Alignment and Detection via 6DoF, Face Pose Estimation ⬇️

We propose real-time, six degrees of freedom (6DoF), 3D face pose estimation without face detection or landmark localization. We observe that estimating the 6DoF rigid transformation of a face is a simpler problem than facial landmark detection, often used for 3D face alignment. In addition, 6DoF offers more information than face bounding box labels. We leverage these observations to make multiple contributions: (a) We describe an easily trained, efficient, Faster R-CNN--based model which regresses 6DoF pose for all faces in the photo, without preliminary face detection. (b) We explain how pose is converted and kept consistent between the input photo and arbitrary crops created while training and evaluating our model. (c) Finally, we show how face poses can replace detection bounding box training labels. Tests on AFLW2000-3D and BIWI show that our method runs at real-time and outperforms state of the art (SotA) face pose estimators. Remarkably, our method also surpasses SotA models of comparable complexity on the WIDER FACE detection benchmark, despite not been optimized on bounding box labels.

3.PePScenes: A Novel Dataset and Baseline for Pedestrian Action Prediction in 3D ⬇️

Predicting the behavior of road users, particularly pedestrians, is vital for safe motion planning in the context of autonomous driving systems. Traditionally, pedestrian behavior prediction has been realized in terms of forecasting future trajectories. However, recent evidence suggests that predicting higher-level actions, such as crossing the road, can help improve trajectory forecasting and planning tasks accordingly. There are a number of existing datasets that cater to the development of pedestrian action prediction algorithms, however, they lack certain characteristics, such as bird's eye view semantic map information, 3D locations of objects in the scene, etc., which are crucial in the autonomous driving context. To this end, we propose a new pedestrian action prediction dataset created by adding per-frame 2D/3D bounding box and behavioral annotations to the popular autonomous driving dataset, nuScenes. In addition, we propose a hybrid neural network architecture that incorporates various data modalities for predicting pedestrian crossing action. By evaluating our model on the newly proposed dataset, the contribution of different data modalities to the prediction task is revealed. The dataset is available at this https URL.

4.Digital rock reconstruction with user-defined properties using conditional generative adversarial networks ⬇️

Uncertainty is ubiquitous with flow in subsurface rocks because of their inherent heterogeneity and lack of in-situ measurements. To complete uncertainty analysis in a multi-scale manner, it is a prerequisite to provide sufficient rock samples. Even though the advent of digital rock technology offers opportunities to reproduce rocks, it still cannot be utilized to provide massive samples due to its high cost, thus leading to the development of diversified mathematical methods. Among them, two-point statistics (TPS) and multi-point statistics (MPS) are commonly utilized, which feature incorporating low-order and high-order statistical information, respectively. Recently, generative adversarial networks (GANs) are becoming increasingly popular since they can reproduce training images with excellent visual and consequent geologic realism. However, standard GANs can only incorporate information from data, while leaving no interface for user-defined properties, and thus may limit the diversity of reconstructed samples. In this study, we propose conditional GANs for digital rock reconstruction, aiming to reproduce samples not only similar to the real training data, but also satisfying user-specified properties. In fact, the proposed framework can realize the targets of MPS and TPS simultaneously by incorporating high-order information directly from rock images with the GANs scheme, while preserving low-order counterparts through conditioning. We conduct three reconstruction experiments, and the results demonstrate that rock type, rock porosity, and correlation length can be successfully conditioned to affect the reconstructed rock images. Furthermore, in contrast to existing GANs, the proposed conditioning enables learning of multiple rock types simultaneously, and thus invisibly saves the computational cost.

5.Improving Panoptic Segmentation at All Scales ⬇️

Crop-based training strategies decouple training resolution from GPU memory consumption, allowing the use of large-capacity panoptic segmentation networks on multi-megapixel images. Using crops, however, can introduce a bias towards truncating or missing large objects. To address this, we propose a novel crop-aware bounding box regression loss (CABB loss), which promotes predictions to be consistent with the visible parts of the cropped objects, while not over-penalizing them for extending outside of the crop. We further introduce a novel data sampling and augmentation strategy which improves generalization across scales by counteracting the imbalanced distribution of object sizes. Combining these two contributions with a carefully designed, top-down panoptic segmentation architecture, we obtain new state-of-the-art results on the challenging Mapillary Vistas (MVD), Indian Driving and Cityscapes datasets, surpassing the previously best approach on MVD by +4.5% PQ and +5.2% mAP.

6.High-resolution global irrigation prediction with Sentinel-2 30m data ⬇️

An accurate and precise understanding of global irrigation usage is crucial for a variety of climate science efforts. Irrigation is highly energy-intensive, and as population growth continues at its current pace, increases in crop need and water usage will have an impact on climate change. Precise irrigation data can help with monitoring water usage and optimizing agricultural yield, particularly in developing countries. Irrigation data, in tandem with precipitation data, can be used to predict water budgets as well as climate and weather modeling. With our research, we produce an irrigation prediction model that combines unsupervised clustering of Normalized Difference Vegetation Index (NDVI) temporal signatures with a precipitation heuristic to label the months that irrigation peaks for each cropland cluster in a given year. We have developed a novel irrigation model and Python package ("Irrigation30") to generate 30m resolution irrigation predictions of cropland worldwide. With a small crowdsourced test set of cropland coordinates and irrigation labels, using a fraction of the resources used by the state-of-the-art NASA-funded GFSAD30 project with irrigation data limited to India and Australia, our model was able to achieve consistency scores in excess of 97% and an accuracy of 92% in a small geo-diverse randomly sampled test set.

7.Lips Don't Lie: A Generalisable and Robust Approach to Face Forgery Detection ⬇️

Although current deep learning-based face forgery detectors achieve impressive performance in constrained scenarios, they are vulnerable to samples created by unseen manipulation methods. Some recent works show improvements in generalisation but rely on cues that are easily corrupted by common post-processing operations such as compression. In this paper, we propose LipForensics, a detection approach capable of both generalising to novel manipulations and withstanding various distortions. LipForensics targets high-level semantic irregularities in mouth movements, which are common in many generated videos. It consists in first pretraining a spatio-temporal network to perform visual speech recognition (lipreading), thus learning rich internal representations related to natural mouth motion. A temporal network is subsequently finetuned on fixed mouth embeddings of real and forged data in order to detect fake videos based on mouth movements without overfitting to low-level, manipulation-specific artefacts. Extensive experiments show that this simple approach significantly surpasses the state-of-the-art in terms of generalisation to unseen manipulations and robustness to perturbations, as well as shed light on the factors responsible for its performance.

8.Deep Neural Networks for COVID-19 Detection and Diagnosis using Images and Acoustic-based Techniques: A Recent Review ⬇️

The new coronavirus disease (COVID-19) has been declared a pandemic since March 2020 by the World Health Organization. It consists of an emerging viral infection with respiratory tropism that could develop atypical pneumonia. Experts emphasize the importance of early detection of those who have the COVID-19 virus. In this way, patients will be isolated from other people and the spread of the virus can be prevented. For this reason, it has become an area of interest to develop early diagnosis and detection methods to ensure a rapid treatment process and prevent the virus from spreading. Since the standard testing system is time-consuming and not available for everyone, alternative early-screening techniques have become an urgent need. In this study, the approaches used in the detection of COVID-19 based on deep learning (DL) algorithms, which have been popular in recent years, have been comprehensively discussed. The advantages and disadvantages of different approaches used in literature are examined in detail. The Computed Tomography of the chest and X-ray images give a rich representation of the patient's lung that is less time-consuming and allows an efficient viral pneumonia detection using the DL algorithms. The first step is the pre-processing of these images to remove noise. Next, deep features are extracted using multiple types of deep models (pre-trained models, generative models, generic neural networks, etc). Finally, the classification is performed using the obtained features to decide whether the patient is infected by coronavirus or it is another lung disease. In this study, we also give a brief review of the latest applications of cough analysis to early screen the COVID-19, and human mobility estimation to limit its spread.

9.ProLab: perceptually uniform projective colour coordinate system ⬇️

In this work, we propose proLab: a new colour coordinate system derived as a 3D projective transformation of CIE XYZ. We show that proLab is far ahead of the widely used CIELAB coordinate system (though inferior to the modern CAM16-UCS) according to perceptual uniformity evaluated by the STRESS metric in reference to the CIEDE2000 colour difference formula. At the same time, angular errors of chromaticity estimation that are standard for linear colour spaces can also be used in proLab since projective transformations preserve the linearity of manifolds. Unlike in linear spaces, angular errors for different hues are normalized according to human colour discrimination thresholds within proLab. We also demonstrate that shot noise in proLab is more homoscedastic than in CAM16-UCS or other standard colour spaces. This makes proLab a convenient coordinate system in which to perform linear colour analysis.

10.Decoupled Self Attention for Accurate One Stage Object Detection ⬇️

As the scale of object detection dataset is smaller than that of image recognition dataset ImageNet, transfer learning has become a basic training method for deep learning object detection models, which will pretrain the backbone network of object detection model on ImageNet dataset to extract features for classification and localization subtasks. However, the classification task focuses on the salient region features of object, while the location task focuses on the edge features of object, so there is certain deviation between the features extracted by pretrained backbone network and the features used for localization task. In order to solve this problem, a decoupled self attention(DSA) module is proposed for one stage object detection models in this paper. DSA includes two decoupled self-attention branches, so it can extract appropriate features for different tasks. It is located between FPN and head networks of subtasks, so it is used to extract global features based on FPN fused features for different tasks independently. Although the network of DSA module is simple, but it can effectively improve the performance of object detection, also it can be easily embedded in many detection models. Our experiments are based on the representative one-stage detection model RetinaNet. In COCO dataset, when ResNet50 and ResNet101 are used as backbone networks, the detection performances can be increased by 0.4% AP and 0.5% AP respectively. When DSA module and object confidence task are applied in RetinaNet together, the detection performances based on ResNet50 and ResNet101 can be increased by 1.0% AP and 1.4% AP respectively. The experiment results show the effectiveness of DSA module.

11.Understanding Image Retrieval Re-Ranking: A Graph Neural Network Perspective ⬇️

The re-ranking approach leverages high-confidence retrieved samples to refine retrieval results, which have been widely adopted as a post-processing tool for image retrieval tasks. However, we notice one main flaw of re-ranking, i.e., high computational complexity, which leads to an unaffordable time cost for real-world applications. In this paper, we revisit re-ranking and demonstrate that re-ranking can be reformulated as a high-parallelism Graph Neural Network (GNN) function. In particular, we divide the conventional re-ranking process into two phases, i.e., retrieving high-quality gallery samples and updating features. We argue that the first phase equals building the k-nearest neighbor graph, while the second phase can be viewed as spreading the message within the graph. In practice, GNN only needs to concern vertices with the connected edges. Since the graph is sparse, we can efficiently update the vertex features. On the Market-1501 dataset, we accelerate the re-ranking processing from 89.2s to 9.4ms with one K40m GPU, facilitating the real-time post-processing. Similarly, we observe that our method achieves comparable or even better retrieval results on the other four image retrieval benchmarks, i.e., VeRi-776, Oxford-5k, Paris-6k and University-1652, with limited time cost. Our code is publicly available.

12.WDNet: Watermark-Decomposition Network for Visible Watermark Removal ⬇️

Visible watermarks are widely-used in images to protect copyright ownership. Analyzing watermark removal helps to reinforce the anti-attack techniques in an adversarial way. Current removal methods normally leverage image-to-image translation techniques. Nevertheless, the uncertainty of the size, shape, color and transparency of the watermarks set a huge barrier for these methods. To combat this, we combine traditional watermarked image decomposition into a two-stage generator, called Watermark-Decomposition Network (WDNet), where the first stage predicts a rough decomposition from the whole watermarked image and the second stage specifically centers on the watermarked area to refine the removal results. The decomposition formulation enables WDNet to separate watermarks from the images rather than simply removing them. We further show that these separated watermarks can serve as extra nutrients for building a larger training dataset and further improving removal performance. Besides, we construct a large-scale dataset named CLWD, which mainly contains colored watermarks, to fill the vacuum of colored watermark removal dataset. Extensive experiments on the public gray-scale dataset LVW and CLWD consistently show that the proposed WDNet outperforms the state-of-the-art approaches both in accuracy and efficiency.

13.Agglomerative Clustering of Handwritten Numerals to Determine Similarity of Different Languages ⬇️

Handwritten numerals of different languages have various characteristics. Similarities and dissimilarities of the languages can be measured by analyzing the extracted features of the numerals. Handwritten numeral datasets are available and accessible for many renowned languages of different regions. In this paper, several handwritten numeral datasets of different languages are collected. Then they are used to find the similarity among those written languages through determining and comparing the similitude of each handwritten numerals. This will help to find which languages have the same or adjacent parent language. Firstly, a similarity measure of two numeral images is constructed with a Siamese network. Secondly, the similarity of the numeral datasets is determined with the help of the Siamese network and a new random sample with replacement similarity averaging technique. Finally, an agglomerative clustering is done based on the similarities of each dataset. This clustering technique shows some very interesting properties of the datasets. The property focused in this paper is the regional resemblance of the datasets. By analyzing the clusters, it becomes easy to identify which languages are originated from similar regions.

14.FlowMOT: 3D Multi-Object Tracking by Scene Flow Association ⬇️

Most end-to-end Multi-Object Tracking (MOT) methods face the problems of low accuracy and poor generalization ability. Although traditional filter-based methods can achieve better results, they are difficult to be endowed with optimal hyperparameters and often fail in varying scenarios. To alleviate these drawbacks, we propose a LiDAR-based 3D MOT framework named FlowMOT, which integrates point-wise motion information into the traditional matching algorithm, enhancing the robustness of the data association. We firstly utilize a scene flow estimation network to obtain implicit motion information between two adjacent frames and calculate the predicted detection for each old tracklet in the previous frame. Then we use Hungarian algorithm to generate optimal matching relations with the ID propagation strategy to finish the tracking task. Experiments on KITTI MOT dataset show that our approach outperforms recent end-to-end methods and achieves competitive performance with the state-of-the-art filter-based method. In addition, ours can work steadily in the various-speed scenes where the filter-based methods may fail.

15.Temporal Relational Modeling with Self-Supervision for Action Segmentation ⬇️

Temporal relational modeling in video is essential for human action understanding, such as action recognition and action segmentation. Although Graph Convolution Networks (GCNs) have shown promising advantages in relation reasoning on many tasks, it is still a challenge to apply graph convolution networks on long video sequences effectively. The main reason is that large number of nodes (i.e., video frames) makes GCNs hard to capture and model temporal relations in videos. To tackle this problem, in this paper, we introduce an effective GCN module, Dilated Temporal Graph Reasoning Module (DTGRM), designed to model temporal relations and dependencies between video frames at various time spans. In particular, we capture and model temporal relations via constructing multi-level dilated temporal graphs where the nodes represent frames from different moments in video. Moreover, to enhance temporal reasoning ability of the proposed model, an auxiliary self-supervised task is proposed to encourage the dilated temporal graph reasoning module to find and correct wrong temporal relations in videos. Our DTGRM model outperforms state-of-the-art action segmentation models on three challenging datasets: 50Salads, Georgia Tech Egocentric Activities (GTEA), and the Breakfast dataset. The code is available at this https URL.

16.Improving Video Instance Segmentation by Light-weight Temporal Uncertainty Estimates ⬇️

Instance segmentation with neural networks is an essential task in environment perception. However, the networks can predict false positive instances with high confidence values and true positives with low ones. Hence, it is important to accurately model the uncertainties of neural networks to prevent safety issues and foster interpretability. In applications such as automated driving the detection of road users like vehicles and pedestrians is of highest interest. We present a temporal approach to detect false positives and investigate uncertainties of instance segmentation networks. Since image sequences are available for online applications, we track instances over multiple frames and create temporal instance-wise aggregated metrics of uncertainty. The prediction quality is estimated by predicting the intersection over union as performance measure. Furthermore, we show how to use uncertainty information to replace the traditional score value from object detection and improve the overall performance of instance segmentation networks.

17.Sign-Agnostic Implicit Learning of Surface Self-Similarities for Shape Modeling and Reconstruction from Raw Point Clouds ⬇️

Shape modeling and reconstruction from raw point clouds of objects stand as a fundamental challenge in vision and graphics research. Classical methods consider analytic shape priors; however, their performance degraded when the scanned points deviate from the ideal conditions of cleanness and completeness. Important progress has been recently made by data-driven approaches, which learn global and/or local models of implicit surface representations from auxiliary sets of training shapes. Motivated from a universal phenomenon that self-similar shape patterns of local surface patches repeat across the entire surface of an object, we aim to push forward the data-driven strategies and propose to learn a local implicit surface network for a shared, adaptive modeling of the entire surface for a direct surface reconstruction from raw point cloud; we also enhance the leveraging of surface self-similarities by improving correlations among the optimized latent codes of individual surface patches. Given that orientations of raw points could be unavailable or noisy, we extend sign agnostic learning into our local implicit model, which enables our recovery of signed implicit fields of local surfaces from the unsigned inputs. We term our framework as Sign-Agnostic Implicit Learning of Surface Self-Similarities (SAIL-S3). With a global post-optimization of local sign flipping, SAIL-S3 is able to directly model raw, un-oriented point clouds and reconstruct high-quality object surfaces. Experiments show its superiority over existing methods.

18.Deep Learning for Material recognition: most recent advances and open challenges ⬇️

Recognizing material from color images is still a challenging problem today. While deep neural networks provide very good results on object recognition and has been the topic of a huge amount of papers in the last decade, their adaptation to material images still requires some works to reach equivalent accuracies. Nevertheless, recent studies achieve very good results in material recognition with deep learning and we propose, in this paper, to review most of them by focusing on three aspects: material image datasets, influence of the context and ad hoc descriptors for material appearance. Every aspect is introduced by a systematic manner and results from representative works are cited. We also present our own studies in this area and point out some open challenges for future works.

19.Scaling Semantic Segmentation Beyond 1K Classes on a Single GPU ⬇️

The state-of-the-art object detection and image classification methods can perform impressively on more than 9k and 10k classes, respectively. In contrast, the number of classes in semantic segmentation datasets is relatively limited. This is not surprising when the restrictions caused by the lack of labeled data and high computation demand for segmentation are considered. In this paper, we propose a novel training methodology to train and scale the existing semantic segmentation models for a large number of semantic classes without increasing the memory overhead. In our embedding-based scalable segmentation approach, we reduce the space complexity of the segmentation model's output from O(C) to O(1), propose an approximation method for ground-truth class probability, and use it to compute cross-entropy loss. The proposed approach is general and can be adopted by any state-of-the-art segmentation model to gracefully scale it for any number of semantic classes with only one GPU. Our approach achieves similar, and in some cases, even better mIoU for Cityscapes, Pascal VOC, ADE20k, COCO-Stuff10k datasets when adopted to DeeplabV3+ model with different backbones. We demonstrate a clear benefit of our approach on a dataset with 1284 classes, bootstrapped from LVIS and COCO annotations, with three times better mIoU than the DeeplabV3+ model.

20.Aggregative Self-Supervised Feature Learning ⬇️

Self-supervised learning (SSL) is an efficient approach that addresses the issue of annotation shortage. The key part in SSL is its proxy task that defines the supervisory signals and drives the learning toward effective feature representations. However, most SSL approaches usually focus on a single proxy task, which greatly limits the expressive power of the learned features and therefore deteriorates the network generalization capacity. In this regard, we hereby propose three strategies of aggregation in terms of complementarity of various forms to boost the robustness of self-supervised learned features. In spatial context aggregative SSL, we contribute a heuristic SSL method that integrates two ad-hoc proxy tasks with spatial context complementarity, modeling global and local contextual features, respectively. We then propose a principled framework of multi-task aggregative self-supervised learning to form a unified representation, with an intent of exploiting feature complementarity among different tasks. Finally, in self-aggregative SSL, we propose to self-complement an existing proxy task with an auxiliary loss function based on a linear centered kernel alignment metric, which explicitly promotes the exploring of where are uncovered by the features learned from a proxy task at hand to further boost the modeling capability. Our extensive experiments on 2D natural image and 3D medical image classification tasks under limited annotation scenarios confirm that the proposed aggregation strategies successfully boost the classification accuracy.

21.Learned Video Codec with Enriched Reconstruction for CLIC P-frame Coding ⬇️

This paper proposes a learning-based video codec, specifically used for Challenge on Learned Image Compression (CLIC, CVPRWorkshop) 2020 P-frame coding. More specifically, we designed a compressor network with Refine-Net for coding residual signals and motion vectors. Also, for motion estimation, we introduced a hierarchical, attention-based ME-Net. To verify our design, we conducted an extensive ablation study on our modules and different input formats. Our video codec demonstrates its performance by using the perfect reference frame at the decoder side specified by the CLIC P-frame Challenge. The experimental result shows that our proposed codec is very competitive with the Challenge top performers in terms of quality metrics.

22.Pyramid-Focus-Augmentation: Medical Image Segmentation with Step-Wise Focus ⬇️

Segmentation of findings in the gastrointestinal tract is a challenging but also an important task which is an important building stone for sufficient automatic decision support systems. In this work, we present our solution for the Medico 2020 task, which focused on the problem of colon polyp segmentation. We present our simple but efficient idea of using an augmentation method that uses grids in a pyramid-like manner (large to small) for segmentation. Our results show that the proposed methods work as indented and can also lead to comparable results when competing with other methods.

23.DSM Refinement with Deep Encoder-Decoder Networks ⬇️

3D city models can be generated from aerial images. However, the calculated DSMs suffer from noise, artefacts, and data holes that have to be manually cleaned up in a time-consuming process. This work presents an approach that automatically refines such DSMs. The key idea is to teach a neural network the characteristics of urban area from reference data. In order to achieve this goal, a loss function consisting of an L1 norm and a feature loss is proposed. These features are constructed using a pre-trained image classification network. To learn to update the height maps, the network architecture is set up based on the concept of deep residual learning and an encoder-decoder structure. The results show that this combination is highly effective in preserving the relevant geometric structures while removing the undesired artefacts and noise.

24.One-Shot Learning with Triplet Loss for Vegetation Classification Tasks ⬇️

Triplet loss function is one of the options that can significantly improve the accuracy of the One-shot Learning tasks. Starting from 2015, many projects use Siamese networks and this kind of loss for face recognition and object classification. In our research, we focused on two tasks related to vegetation. The first one is plant disease detection on 25 classes of five crops (grape, cotton, wheat, cucumbers, and corn). This task is motivated because harvest losses due to diseases is a serious problem for both large farming structures and rural families. The second task is the identification of moss species (5 classes). Mosses are natural bioaccumulators of pollutants; therefore, they are used in environmental monitoring programs. The identification of moss species is an important step in the sample preprocessing. In both tasks, we used self-collected image databases. We tried several deep learning architectures and approaches. Our Siamese network architecture with a triplet loss function and MobileNetV2 as a base network showed the most impressive results in both above-mentioned tasks. The average accuracy for plant disease detection amounted to over 97.8% and 97.6% for moss species classification.

25.HR-Depth: High Resolution Self-Supervised Monocular Depth Estimation ⬇️

Self-supervised learning shows great potential in monoculardepth estimation, using image sequences as the only source ofsupervision. Although people try to use the high-resolutionimage for depth estimation, the accuracy of prediction hasnot been significantly improved. In this work, we find thecore reason comes from the inaccurate depth estimation inlarge gradient regions, making the bilinear interpolation er-ror gradually disappear as the resolution increases. To obtainmore accurate depth estimation in large gradient regions, itis necessary to obtain high-resolution features with spatialand semantic information. Therefore, we present an improvedDepthNet, HR-Depth, with two effective strategies: (1) re-design the skip-connection in DepthNet to get better high-resolution features and (2) propose feature fusion Squeeze-and-Excitation(fSE) module to fuse feature more efficiently.Using Resnet-18 as the encoder, HR-Depth surpasses all pre-vious state-of-the-art(SoTA) methods with the least param-eters at both high and low resolution. Moreover, previousstate-of-the-art methods are based on fairly complex and deepnetworks with a mass of parameters which limits their realapplications. Thus we also construct a lightweight networkwhich uses MobileNetV3 as encoder. Experiments show thatthe lightweight network can perform on par with many largemodels like Monodepth2 at high-resolution with only20%parameters. All codes and models will be available at this https URL.

26.The Open Brands Dataset: Unified brand detection and recognition at scale ⬇️

Intellectual property protection(IPP) have received more and more attention recently due to the development of the global e-commerce platforms. brand recognition plays a significant role in IPP. Recent studies for brand recognition and detection are based on small-scale datasets that are not comprehensive enough when exploring emerging deep learning techniques. Moreover, it is challenging to evaluate the true performance of brand detection methods in realistic and open scenes. In order to tackle these problems, we first define the special issues of brand detection and recognition compared with generic object detection. Second, a novel brands benchmark called "Open Brands" is established. The dataset contains 1,437,812 images which have brands and 50,000 images without any brand. The part with brands in Open Brands contains 3,113,828 instances annotated in 3 dimensions: 4 types, 559 brands and 1216 logos. To the best of our knowledge, it is the largest dataset for brand detection and recognition with rich annotations. We provide in-depth comprehensive statistics about the dataset, validate the quality of the annotations and study how the performance of many modern models evolves with an increasing amount of training data. Third, we design a network called "Brand Net" to handle brand recognition. Brand Net gets state-of-art mAP on Open Brand compared with existing detection methods.

27.Articulated Shape Matching Using Laplacian Eigenfunctions and Unsupervised Point Registration ⬇️

Matching articulated shapes represented by voxel-sets reduces to maximal sub-graph isomorphism when each set is described by a weighted graph. Spectral graph theory can be used to map these graphs onto lower dimensional spaces and match shapes by aligning their embeddings in virtue of their invariance to change of pose. Classical graph isomorphism schemes relying on the ordering of the eigenvalues to align the eigenspaces fail when handling large data-sets or noisy data. We derive a new formulation that finds the best alignment between two congruent $K$-dimensional sets of points by selecting the best subset of eigenfunctions of the Laplacian matrix. The selection is done by matching eigenfunction signatures built with histograms, and the retained set provides a smart initialization for the alignment problem with a considerable impact on the overall performance. Dense shape matching casted into graph matching reduces then, to point registration of embeddings under orthogonal transformations; the registration is solved using the framework of unsupervised clustering and the EM algorithm. Maximal subset matching of non identical shapes is handled by defining an appropriate outlier class. Experimental results on challenging examples show how the algorithm naturally treats changes of topology, shape variations and different sampling densities.

28.Intrinsic Image Captioning Evaluation ⬇️

The image captioning task is about to generate suitable descriptions from images. For this task there can be several challenges such as accuracy, fluency and diversity. However there are few metrics that can cover all these properties while evaluating results of captioning this http URL this paper we first conduct a comprehensive investigation on contemporary metrics. Motivated by the auto-encoder mechanism and the research advances of word embeddings we propose a learning based metrics for image captioning, which we call Intrinsic Image Captioning Evaluation(I2CE). We select several state-of-the-art image captioning models and test their performances on MS COCO dataset with respects to both contemporary metrics and the proposed I2CE. Experiment results show that our proposed method can keep robust performance and give more flexible scores to candidate captions when encountered with semantic similar expression or less aligned semantics. On this concern the proposed metric could serve as a novel indicator on the intrinsic information between captions, which may be complementary to the existing ones.

29.Combining Similarity and Adversarial Learning to Generate Visual Explanation: Application to Medical Image Classification ⬇️

Explaining decisions of black-box classifiers is paramount in sensitive domains such as medical imaging since clinicians confidence is necessary for adoption. Various explanation approaches have been proposed, among which perturbation based approaches are very promising. Within this class of methods, we leverage a learning framework to produce our visual explanations method. From a given classifier, we train two generators to produce from an input image the so called similar and adversarial images. The similar image shall be classified as the input image whereas the adversarial shall not. Visual explanation is built as the difference between these two generated images. Using metrics from the literature, our method outperforms state-of-the-art approaches. The proposed approach is model-agnostic and has a low computation burden at prediction time. Thus, it is adapted for real-time systems. Finally, we show that random geometric augmentations applied to the original image play a regularization role that improves several previously proposed explanation methods. We validate our approach on a large chest X-ray database.

30.Morphology on categorical distributions ⬇️

The categorical distribution is a natural representation of uncertainty in multi-class segmentations. In the two-class case the categorical distribution reduces to the Bernoulli distribution, for which grayscale morphology provides a range of useful operations. In the general case, applying morphological operations on uncertain multi-class segmentations is not straightforward as an image of categorical distributions is not a complete lattice. Although morphology on color images has received wide attention, this is not so for color-coded or categorical images and even less so for images of categorical distributions. In this work, we establish a set of requirements for morphology on categorical distributions by combining classic morphology with a probabilistic view. We then define operators respecting these requirements, introduce protected operations on categorical distributions and illustrate the utility of these operators on two example tasks: modeling annotator bias in brain tumor segmentations and segmenting vesicle instances from the predictions of a multi-class U-Net.

31.Multi Modal Adaptive Normalization for Audio to Video Generation ⬇️

Speech-driven facial video generation has been a complex problem due to its multi-modal aspects namely audio and video domain. The audio comprises lots of underlying features such as expression, pitch, loudness, prosody(speaking style) and facial video has lots of variability in terms of head movement, eye blinks, lip synchronization and movements of various facial action units along with temporal smoothness. Synthesizing highly expressive facial videos from the audio input and static image is still a challenging task for generative adversarial networks. In this paper, we propose a multi-modal adaptive normalization(MAN) based architecture to synthesize a talking person video of arbitrary length using as input: an audio signal and a single image of a person. The architecture uses the multi-modal adaptive normalization, keypoint heatmap predictor, optical flow predictor and class activation map[58] based layers to learn movements of expressive facial components and hence generates a highly expressive talking-head video of the given person. The multi-modal adaptive normalization uses the various features of audio and video such as Mel spectrogram, pitch, energy from audio signals and predicted keypoint heatmap/optical flow and a single image to learn the respective affine parameters to generate highly expressive video. Experimental evaluation demonstrates superior performance of the proposed method as compared to Realistic Speech-Driven Facial Animation with GANs(RSDGAN) [53], Speech2Vid [10], and other approaches, on multiple quantitative metrics including: SSIM (structural similarity index), PSNR (peak signal to noise ratio), CPBD (image sharpness), WER(word error rate), blinks/sec and LMD(landmark distance). Further, qualitative evaluation and Online Turing tests demonstrate the efficacy of our approach.

32.Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer ⬇️

Unsupervised domain adaptation (UDA) aims to transfer knowledge from a related but different well-labeled source domain to a new unlabeled target domain. Most existing UDA methods require access to the source data, and thus are not applicable when the data are confidential and not shareable due to privacy concerns. This paper aims to tackle a realistic setting with only a classification model available trained over, instead of accessing to, the source data. To effectively utilize the source model for adaptation, we propose a novel approach called Source HypOthesis Transfer (SHOT), which learns the feature extraction module for the target domain by fitting the target data features to the frozen source classification module (representing classification hypothesis). Specifically, SHOT exploits both information maximization and self-supervised learning for the feature extraction module learning to ensure the target features are implicitly aligned with the features of unseen source data via the same hypothesis. Furthermore, we propose a new labeling transfer strategy, which separates the target data into two splits based on the confidence of predictions (labeling information), and then employ semi-supervised learning to improve the accuracy of less-confident predictions in the target domain. We denote labeling transfer as SHOT++ if the predictions are obtained by SHOT. Extensive experiments on both digit classification and object recognition tasks show that SHOT and SHOT++ achieve results surpassing or comparable to the state-of-the-arts, demonstrating the effectiveness of our approaches for various visual domain adaptation problems.

33.Learning Category-level Shape Saliency via Deep Implicit Surface Networks ⬇️

This paper is motivated from a fundamental curiosity on what defines a category of object shapes. For example, we may have the common knowledge that a plane has wings, and a chair has legs. Given the large shape variations among different instances of a same category, we are formally interested in developing a quantity defined for individual points on a continuous object surface; the quantity specifies how individual surface points contribute to the formation of the shape as the category. We term such a quantity as category-level shape saliency or shape saliency for short. Technically, we propose to learn saliency maps for shape instances of a same category from a deep implicit surface network; sensible saliency scores for sampled points in the implicit surface field are predicted by constraining the capacity of input latent code. We also enhance the saliency prediction with an additional loss of contrastive training. We expect such learned surface maps of shape saliency to have the properties of smoothness, symmetry, and semantic representativeness. We verify these properties by comparing our method with alternative ways of saliency computation. Notably, we show that by leveraging the learned shape saliency, we are able to reconstruct either category-salient or instance-specific parts of object surfaces; semantic representativeness of the learned saliency is also reflected in its efficacy to guide the selection of surface points for better point cloud classification.

34.Semantic Layout Manipulation with High-Resolution Sparse Attention ⬇️

We tackle the problem of semantic image layout manipulation, which aims to manipulate an input image by editing its semantic label map. A core problem of this task is how to transfer visual details from the input images to the new semantic layout while making the resulting image visually realistic. Recent work on learning cross-domain correspondence has shown promising results for global layout transfer with dense attention-based warping. However, this method tends to lose texture details due to the lack of smoothness and resolution in the correspondence and warped images. To adapt this paradigm for the layout manipulation task, we propose a high-resolution sparse attention module that effectively transfers visual details to new layouts at a resolution up to 512x512. To further improve visual quality, we introduce a novel generator architecture consisting of a semantic encoder and a two-stage decoder for coarse-to-fine synthesis. Experiments on the ADE20k and Places365 datasets demonstrate that our proposed approach achieves substantial improvements over the existing inpainting and layout manipulation methods.

35.Information-Theoretic Segmentation by Inpainting Error Maximization ⬇️

We study image segmentation from an information-theoretic perspective, proposing a novel adversarial method that performs unsupervised segmentation by partitioning images into maximally independent sets. More specifically, we group image pixels into foreground and background, with the goal of minimizing predictability of one set from the other. An easily computed loss drives a greedy search process to maximize inpainting error over these partitions. Our method does not involve training deep networks, is computationally cheap, class-agnostic, and even applicable in isolation to a single unlabeled image. Experiments demonstrate that it achieves a new state-of-the-art in unsupervised segmentation quality, while being substantially faster and more general than competing approaches.

36.TDAF: Top-Down Attention Framework for Vision Tasks ⬇️

Human attention mechanisms often work in a top-down manner, yet it is not well explored in vision research. Here, we propose the Top-Down Attention Framework (TDAF) to capture top-down attentions, which can be easily adopted in most existing models. The designed Recursive Dual-Directional Nested Structure in it forms two sets of orthogonal paths, recursive and structural ones, where bottom-up spatial features and top-down attention features are extracted respectively. Such spatial and attention features are nested deeply, therefore, the proposed framework works in a mixed top-down and bottom-up manner. Empirical evidence shows that our TDAF can capture effective stratified attention information and boost performance. ResNet with TDAF achieves 2.0% improvements on ImageNet. For object detection, the performance is improved by 2.7% AP over FCOS. For pose estimation, TDAF improves the baseline by 1.6%. And for action recognition, the 3D-ResNet adopting TDAF achieves improvements of 1.7% accuracy.

37.Deep Optimized Priors for 3D Shape Modeling and Reconstruction ⬇️

Many learning-based approaches have difficulty scaling to unseen data, as the generality of its learned prior is limited to the scale and variations of the training samples. This holds particularly true with 3D learning tasks, given the sparsity of 3D datasets available. We introduce a new learning framework for 3D modeling and reconstruction that greatly improves the generalization ability of a deep generator. Our approach strives to connect the good ends of both learning-based and optimization-based methods. In particular, unlike the common practice that fixes the pre-trained priors at test time, we propose to further optimize the learned prior and latent code according to the input physical measurements after the training. We show that the proposed strategy effectively breaks the barriers constrained by the pre-trained priors and could lead to high-quality adaptation to unseen data. We realize our framework using the implicit surface representation and validate the efficacy of our approach in a variety of challenging tasks that take highly sparse or collapsed observations as input. Experimental results show that our approach compares favorably with the state-of-the-art methods in terms of both generality and accuracy.

38.INSPIRE: Intensity and Spatial Information-Based Deformable Image Registration ⬇️

We present INSPIRE, a top-performing general-purpose method for deformable image registration. INSPIRE extends our existing symmetric registration framework based on distances combining intensity and spatial information to an elastic B-splines based transformation model. We also present several theoretical and algorithmic improvements which provide high computational efficiency and thereby applicability of the framework in a wide range of real scenarios. We show that the proposed method delivers both highly accurate as well as stable and robust registration results. We evaluate the method on a synthetic dataset created from retinal images, consisting of thin networks of vessels, where INSPIRE exhibits excellent performance, substantially outperforming the reference methods. We also evaluate the method on four benchmark datasets of 3D images of brains, for a total of 2088 pairwise registrations; a comparison with 15 other state-of-the-art methods reveals that INSPIRE provides the best overall performance. Code is available at this http URL.

39.Knowledge-Routed Visual Question Reasoning: Challenges for Deep Representation Embedding ⬇️

Though beneficial for encouraging the Visual Question Answering (VQA) models to discover the underlying knowledge by exploiting the input-output correlation beyond image and text contexts, the existing knowledge VQA datasets are mostly annotated in a crowdsource way, e.g., collecting questions and external reasons from different users via the internet. In addition to the challenge of knowledge reasoning, how to deal with the annotator bias also remains unsolved, which often leads to superficial over-fitted correlations between questions and answers. To address this issue, we propose a novel dataset named Knowledge-Routed Visual Question Reasoning for VQA model evaluation. Considering that a desirable VQA model should correctly perceive the image context, understand the question, and incorporate its learned knowledge, our proposed dataset aims to cutoff the shortcut learning exploited by the current deep embedding models and push the research boundary of the knowledge-based visual question reasoning. Specifically, we generate the question-answer pair based on both the Visual Genome scene graph and an external knowledge base with controlled programs to disentangle the knowledge from other biases. The programs can select one or two triplets from the scene graph or knowledge base to push multi-step reasoning, avoid answer ambiguity, and balanced the answer distribution. In contrast to the existing VQA datasets, we further imply the following two major constraints on the programs to incorporate knowledge reasoning: i) multiple knowledge triplets can be related to the question, but only one knowledge relates to the image object. This can enforce the VQA model to correctly perceive the image instead of guessing the knowledge based on the given question solely; ii) all questions are based on different knowledge, but the candidate answers are the same for both the training and test sets.

40.Meticulous Object Segmentation ⬇️

Compared with common image segmentation tasks targeted at low-resolution images, higher resolution detailed image segmentation receives much less attention. In this paper, we propose and study a task named Meticulous Object Segmentation (MOS), which is focused on segmenting well-defined foreground objects with elaborate shapes in high resolution images (e.g. 2k - 4k). To this end, we propose the MeticulousNet which leverages a dedicated decoder to capture the object boundary details. Specifically, we design a Hierarchical Point-wise Refining (HierPR) block to better delineate object boundaries, and reformulate the decoding process as a recursive coarse to fine refinement of the object mask. To evaluate segmentation quality near object boundaries, we propose the Meticulosity Quality (MQ) score considering both the mask coverage and boundary precision. In addition, we collect a MOS benchmark dataset including 600 high quality images with complex objects. We provide comprehensive empirical evidence showing that MeticulousNet can reveal pixel-accurate segmentation boundaries and is superior to state-of-the-art methods for high resolution object segmentation tasks.

41.Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation ⬇️

Building instance segmentation models that are data-efficient and can handle rare object categories is an important challenge in computer vision. Leveraging data augmentations is a promising direction towards addressing this challenge. Here, we perform a systematic study of the Copy-Paste augmentation ([13, 12]) for instance segmentation where we randomly paste objects onto an image. Prior studies on Copy-Paste relied on modeling the surrounding visual context for pasting the objects. However, we find that the simple mechanism of pasting objects randomly is good enough and can provide solid gains on top of strong baselines. Furthermore, we show Copy-Paste is additive with semi-supervised methods that leverage extra data through pseudo labeling (e.g. self-training). On COCO instance segmentation, we achieve 49.1 mask AP and 57.3 box AP, an improvement of +0.6 mask AP and +1.5 box AP over the previous state-of-the-art. We further demonstrate that Copy-Paste can lead to significant improvements on the LVIS benchmark. Our baseline model outperforms the LVIS 2020 Challenge winning entry by +3.6 mask AP on rare categories.

42.MSAF: Multimodal Split Attention Fusion ⬇️

Multimodal learning mimics the reasoning process of the human multi-sensory system, which is used to perceive the surrounding world. While making a prediction, the human brain tends to relate crucial cues from multiple sources of information. In this work, we propose a novel multimodal fusion module that learns to emphasize more contributive features across all modalities. Specifically, the proposed Multimodal Split Attention Fusion (MSAF) module splits each modality into channel-wise equal feature blocks and creates a joint representation that is used to generate soft attention for each channel across the feature blocks. Further, the MSAF module is designed to be compatible with features of various spatial dimensions and sequence lengths, suitable for both CNNs and RNNs. Thus, MSAF can be easily added to fuse features of any unimodal networks and utilize existing pretrained unimodal model weights. To demonstrate the effectiveness of our fusion module, we design three multimodal networks with MSAF for emotion recognition, sentiment analysis, and action recognition tasks. Our approach achieves competitive results in each task and outperforms other application-specific networks and multimodal fusion benchmarks.

43.FSOCO: The Formula Student Objects in Context Dataset ⬇️

This paper presents the FSOCO dataset, a collaborative dataset for vision-based cone detection systems in Formula Student Driverless competitions. It contains human annotated ground truth labels for both bounding boxes and instance-wise segmentation masks. The data buy-in philosophy of FSOCO asks student teams to contribute to the database first before being granted access ensuring continuous growth. By providing clear labeling guidelines and tools for a sophisticated raw image selection, new annotations are guaranteed to meet the desired quality. The effectiveness of the approach is shown by comparing prediction results of a network trained on FSOCO and its unregulated predecessor. The FSOCO dataset can be found at this http URL.

44.Location-aware Single Image Reflection Removal ⬇️

This paper proposes a novel location-aware deep learning-based single image reflection removal method. Our network has a reflection detection module to regress a probabilistic reflection confidence map, taking multi-scale Laplacian features as inputs. This probabilistic map tells whether a region is reflection-dominated or transmission-dominated. The novelty is that we use the reflection confidence map as the cues for the network to learn how to encode the reflection information adaptively and control the feature flow when predicting reflection and transmission layers. The integration of location information into the network significantly improves the quality of reflection removal results. Besides, a set of learnable Laplacian kernel parameters is introduced to facilitate the extraction of discriminative Laplacian features for reflection detection. We design our network as a recurrent network to progressively refine each iteration's reflection removal results. Extensive experiments verify the superior performance of the proposed method over state-of-the-art approaches.

45.Iterative Knowledge Exchange Between Deep Learning and Space-Time Spectral Clustering for Unsupervised Segmentation in Videos ⬇️

We propose a dual system for unsupervised object segmentation in video, which brings together two modules with complementary properties: a space-time graph that discovers objects in videos and a deep network that learns powerful object features. The system uses an iterative knowledge exchange policy. A novel spectral space-time clustering process on the graph produces unsupervised segmentation masks passed to the network as pseudo-labels. The net learns to segment in single frames what the graph discovers in video and passes back to the graph strong image-level features that improve its node-level features in the next iteration. Knowledge is exchanged for several cycles until convergence. The graph has one node per each video pixel, but the object discovery is fast. It uses a novel power iteration algorithm computing the main space-time cluster as the principal eigenvector of a special Feature-Motion matrix without actually computing the matrix. The thorough experimental analysis validates our theoretical claims and proves the effectiveness of the cyclical knowledge exchange. We also perform experiments on the supervised scenario, incorporating features pretrained with human supervision. We achieve state-of-the-art level on unsupervised and supervised scenarios on four challenging datasets: DAVIS, SegTrack, YouTube-Objects, and DAVSOD.

46.DFR: Deep Feature Reconstruction for Unsupervised Anomaly Segmentation ⬇️

Automatic detecting anomalous regions in images of objects or textures without priors of the anomalies is challenging, especially when the anomalies appear in very small areas of the images, making difficult-to-detect visual variations, such as defects on manufacturing products. This paper proposes an effective unsupervised anomaly segmentation approach that can detect and segment out the anomalies in small and confined regions of images. Concretely, we develop a multi-scale regional feature generator that can generate multiple spatial context-aware representations from pre-trained deep convolutional networks for every subregion of an image. The regional representations not only describe the local characteristics of corresponding regions but also encode their multiple spatial context information, making them discriminative and very beneficial for anomaly detection. Leveraging these descriptive regional features, we then design a deep yet efficient convolutional autoencoder and detect anomalous regions within images via fast feature reconstruction. Our method is simple yet effective and efficient. It advances the state-of-the-art performances on several benchmark datasets and shows great potential for real applications.

47.Learning Heatmap-Style Jigsaw Puzzles Provides Good Pretraining for 2D Human Pose Estimation ⬇️

The target of 2D human pose estimation is to locate the keypoints of body parts from input 2D images. State-of-the-art methods for pose estimation usually construct pixel-wise heatmaps from keypoints as labels for learning convolution neural networks, which are usually initialized randomly or using classification models on ImageNet as their backbones. We note that 2D pose estimation task is highly dependent on the contextual relationship between image patches, thus we introduce a self-supervised method for pretraining 2D pose estimation networks. Specifically, we propose Heatmap-Style Jigsaw Puzzles (HSJP) problem as our pretext-task, whose target is to learn the location of each patch from an image composed of shuffled patches. During our pretraining process, we only use images of person instances in MS-COCO, rather than introducing extra and much larger ImageNet dataset. A heatmap-style label for patch location is designed and our learning process is in a non-contrastive way. The weights learned by HSJP pretext task are utilised as backbones of 2D human pose estimator, which are then finetuned on MS-COCO human keypoints dataset. With two popular and strong 2D human pose estimators, HRNet and SimpleBaseline, we evaluate mAP score on both MS-COCO validation and test-dev datasets. Our experiments show that downstream pose estimators with our self-supervised pretraining obtain much better performance than those trained from scratch, and are comparable to those using ImageNet classification models as their initial backbones.

48.MSVD-Turkish: A Comprehensive Multimodal Dataset for Integrated Vision and Language Research in Turkish ⬇️

Automatic generation of video descriptions in natural language, also called video captioning, aims to understand the visual content of the video and produce a natural language sentence depicting the objects and actions in the scene. This challenging integrated vision and language problem, however, has been predominantly addressed for English. The lack of data and the linguistic properties of other languages limit the success of existing approaches for such languages. In this paper we target Turkish, a morphologically rich and agglutinative language that has very different properties compared to English. To do so, we create the first large scale video captioning dataset for this language by carefully translating the English descriptions of the videos in the MSVD (Microsoft Research Video Description Corpus) dataset into Turkish. In addition to enabling research in video captioning in Turkish, the parallel English-Turkish descriptions also enables the study of the role of video context in (multimodal) machine translation. In our experiments, we build models for both video captioning and multimodal machine translation and investigate the effect of different word segmentation approaches and different neural architectures to better address the properties of Turkish. We hope that the MSVD-Turkish dataset and the results reported in this work will lead to better video captioning and multimodal machine translation models for Turkish and other morphology rich and agglutinative languages.

49.EfficientPose: Efficient Human Pose Estimation with Neural Architecture Search ⬇️

Human pose estimation from image and video is a vital task in many multimedia applications. Previous methods achieve great performance but rarely take efficiency into consideration, which makes it difficult to implement the networks on resource-constrained devices. Nowadays real-time multimedia applications call for more efficient models for better interactions. Moreover, most deep neural networks for pose estimation directly reuse the networks designed for image classification as the backbone, which are not yet optimized for the pose estimation task. In this paper, we propose an efficient framework targeted at human pose estimation including two parts, the efficient backbone and the efficient head. By implementing the differentiable neural architecture search method, we customize the backbone network design for pose estimation and reduce the computation cost with negligible accuracy degradation. For the efficient head, we slim the transposed convolutions and propose a spatial information correction module to promote the performance of the final prediction. In experiments, we evaluate our networks on the MPII and COCO datasets. Our smallest model has only 0.65 GFLOPs with 88.1% PCKh@0.5 on MPII and our large model has only 2 GFLOPs while its accuracy is competitive with the state-of-the-art large model, i.e., HRNet with 9.5 GFLOPs.

50.Robust Real-Time Pedestrian Detection on Embedded Devices ⬇️

Detection of pedestrians on embedded devices, such as those on-board of robots and drones, has many applications including road intersection monitoring, security, crowd monitoring and surveillance, to name a few. However, the problem can be challenging due to continuously-changing camera viewpoint and varying object appearances as well as the need for lightweight algorithms suitable for embedded systems. This paper proposes a robust framework for pedestrian detection in many footages. The framework performs fine and coarse detections on different image regions and exploits temporal and spatial characteristics to attain enhanced accuracy and real time performance on embedded boards. The framework uses the Yolo-v3 object detection [1] as its backbone detector and runs on the Nvidia Jetson TX2 embedded board, however other detectors and/or boards can be used as well. The performance of the framework is demonstrated on two established datasets and its achievement of the second place in CVPR 2019 Embedded Real-Time Inference (ERTI) Challenge.

51.Improving Image Captioning by Leveraging Intra- and Inter-layer Global Representation in Transformer Network ⬇️

Transformer-based architectures have shown great success in image captioning, where object regions are encoded and then attended into the vectorial representations to guide the caption decoding. However, such vectorial representations only contain region-level information without considering the global information reflecting the entire image, which fails to expand the capability of complex multi-modal reasoning in image captioning. In this paper, we introduce a Global Enhanced Transformer (termed GET) to enable the extraction of a more comprehensive global representation, and then adaptively guide the decoder to generate high-quality captions. In GET, a Global Enhanced Encoder is designed for the embedding of the global feature, and a Global Adaptive Decoder are designed for the guidance of the caption generation. The former models intra- and inter-layer global representation by taking advantage of the proposed Global Enhanced Attention and a layer-wise fusion module. The latter contains a Global Adaptive Controller that can adaptively fuse the global information into the decoder to guide the caption generation. Extensive experiments on MS COCO dataset demonstrate the superiority of our GET over many state-of-the-arts.

52.PoNA: Pose-guided Non-local Attention for Human Pose Transfer ⬇️

Human pose transfer, which aims at transferring the appearance of a given person to a target pose, is very challenging and important in many applications. Previous work ignores the guidance of pose features or only uses local attention mechanism, leading to implausible and blurry results. We propose a new human pose transfer method using a generative adversarial network (GAN) with simplified cascaded blocks. In each block, we propose a pose-guided non-local attention (PoNA) mechanism with a long-range dependency scheme to select more important regions of image features to transfer. We also design pre-posed image-guided pose feature update and post-posed pose-guided image feature update to better utilize the pose and image features. Our network is simple, stable, and easy to train. Quantitative and qualitative results on Market-1501 and DeepFashion datasets show the efficacy and efficiency of our model. Compared with state-of-the-art methods, our model generates sharper and more realistic images with rich details, while having fewer parameters and faster speed. Furthermore, our generated images can help to alleviate data insufficiency for person re-identification.

53.One-Shot Object Localization in Medical Images based on Relative Position Regression ⬇️

Deep learning networks have shown promising performance for accurate object localization in medial images, but require large amount of annotated data for supervised training, which is expensive and expertise burdensome. To address this problem, we present a one-shot framework for organ and landmark localization in volumetric medical images, which does not need any annotation during the training stage and could be employed to locate any landmarks or organs in test images given a support (reference) image during the inference stage. Our main idea comes from that tissues and organs from different human bodies have a similar relative position and context. Therefore, we could predict the relative positions of their non-local patches, thus locate the target organ. Our framework is composed of three parts: (1) A projection network trained to predict the 3D offset between any two patches from the same volume, where human annotations are not required. In the inference stage, it takes one given landmark in a reference image as a support patch and predicts the offset from a random patch to the corresponding landmark in the test (query) volume. (2) A coarse-to-fine framework contains two projection networks, providing more accurate localization of the target. (3) Based on the coarse-to-fine model, we transfer the organ boundingbox (B-box) detection to locating six extreme points along x, y and z directions in the query volume. Experiments on multi-organ localization from head-and-neck (HaN) CT volumes showed that our method acquired competitive performance in real time, which is more accurate and 10^5 times faster than template matching methods with the same setting. Code is available: this https URL.

54.Semi-supervised Segmentation via Uncertainty Rectified Pyramid Consistency and Its Application to Gross Target Volume of Nasopharyngeal Carcinoma ⬇️

Gross Target Volume (GTV) segmentation plays an irreplaceable role in radiotherapy planning for Nasopharyngeal Carcinoma (NPC). Despite that convolutional neural networks (CNN) have achieved good performance for this task, they rely on a large set of labeled images for training, which is expensive and time-consuming to acquire. Recently, semi-supervised methods that learn from a small set of labeled images with a large set of unlabeled images have shown potential for dealing with this problem, but it is still challenging to train a high-performance model with the limited number of labeled data. In this paper, we propose a novel framework with Uncertainty Rectified Pyramid Consistency (URPC) regularization for semi-supervised NPC GTV segmentation. Concretely, we extend a backbone segmentation network to produce pyramid predictions at different scales, the pyramid predictions network (PPNet) was supervised by the ground truth of labeled images and a multi-scale consistency loss for unlabeled images, motivated by the fact that prediction at different scales for the same input should be similar and consistent. However, due to the different resolution of these predictions, encouraging them to be consistent at each pixel directly is not robust and may bring much noise and lead to a performance drop. To deal with this dilemma, we further design a novel uncertainty rectifying module to enable the framework to gradually learn from meaningful and reliable consensual regions at different scales. Extensive experiments on our collected NPC dataset with 258 volumes show that our method can largely improve performance by incorporating the unlabeled data, and this framework achieves a promising result compared with existing semi-supervised methods, which achieves 81.22% of mean DSC and 1.88 voxels of mean ASD on the test set, where the only 20% of the training set were annotated.

55.Uncertainty Estimation in Deep Neural Networks for Point Cloud Segmentation in Factory Planning ⬇️

The digital factory provides undoubtedly a great potential for future production systems in terms of efficiency and effectivity. A key aspect on the way to realize the digital copy of a real factory is the understanding of complex indoor environments on the basis of 3D data. In order to generate an accurate factory model including the major components, i.e. building parts, product assets and process details, the 3D data collected during digitalization can be processed with advanced methods of deep learning. In this work, we propose a fully Bayesian and an approximate Bayesian neural network for point cloud segmentation. This allows us to analyze how different ways of estimating uncertainty in these networks improve segmentation results on raw 3D point clouds. We achieve superior model performance for both, the Bayesian and the approximate Bayesian model compared to the frequentist one. This performance difference becomes even more striking when incorporating the networks' uncertainty in their predictions. For evaluation we use the scientific data set S3DIS as well as a data set, which was collected by the authors at a German automotive production plant. The methods proposed in this work lead to more accurate segmentation results and the incorporation of uncertainty information makes this approach especially applicable to safety critical applications.

56.Efficient Human Pose Estimation by Learning Deeply Aggregated Representations ⬇️

In this paper, we propose an efficient human pose estimation network (DANet) by learning deeply aggregated representations. Most existing models explore multi-scale information mainly from features with different spatial sizes. Powerful multi-scale representations usually rely on the cascaded pyramid framework. This framework largely boosts the performance but in the meanwhile makes networks very deep and complex. Instead, we focus on exploiting multi-scale information from layers with different receptive-field sizes and then making full of use this information by improving the fusion method. Specifically, we propose an orthogonal attention block (OAB) and a second-order fusion unit (SFU). The OAB learns multi-scale information from different layers and enhances them by encouraging them to be diverse. The SFU adaptively selects and fuses diverse multi-scale information and suppress the redundant ones. This could maximize the effective information in final fused representations. With the help of OAB and SFU, our single pyramid network may be able to generate deeply aggregated representations that contain even richer multi-scale information and have a larger representing capacity than that of cascaded networks. Thus, our networks could achieve comparable or even better accuracy with much smaller model complexity. Specifically, our \mbox{DANet-72} achieves $70.5$ in AP score on COCO test-dev set with only $1.0G$ FLOPs. Its speed on a CPU platform achieves $58$ Persons-Per-Second~(PPS).

57.Split then Refine: Stacked Attention-guided ResUNets for Blind Single Image Visible Watermark Removal ⬇️

Digital watermark is a commonly used technique to protect the copyright of medias. Simultaneously, to increase the robustness of watermark, attacking technique, such as watermark removal, also gets the attention from the community. Previous watermark removal methods require to gain the watermark location from users or train a multi-task network to recover the background indiscriminately. However, when jointly learning, the network performs better on watermark detection than recovering the texture. Inspired by this observation and to erase the visible watermarks blindly, we propose a novel two-stage framework with a stacked attention-guided ResUNets to simulate the process of detection, removal and refinement. In the first stage, we design a multi-task network called SplitNet. It learns the basis features for three sub-tasks altogether while the task-specific features separately use multiple channel attentions. Then, with the predicted mask and coarser restored image, we design RefineNet to smooth the watermarked region with a mask-guided spatial attention. Besides network structure, the proposed algorithm also combines multiple perceptual losses for better quality both visually and numerically. We extensively evaluate our algorithm over four different datasets under various settings and the experiments show that our approach outperforms other state-of-the-art methods by a large margin. The code is available at this http URL.

58.Effective multi-view registration of point sets based on student's t mixture model ⬇️

Recently, Expectation-maximization (EM) algorithm has been introduced as an effective means to solve multi-view registration problem. Most of the previous methods assume that each data point is drawn from the Gaussian Mixture Model (GMM), which is difficult to deal with the noise with heavy-tail or outliers. Accordingly, this paper proposed an effective registration method based on Student's t Mixture Model (StMM). More specially, we assume that each data point is drawn from one unique StMM, where its nearest neighbors (NNs) in other point sets are regarded as the t-distribution centroids with equal covariances, membership probabilities, and fixed degrees of freedom. Based on this assumption, the multi-view registration problem is formulated into the maximization of the likelihood function including all rigid transformations. Subsequently, the EM algorithm is utilized to optimize rigid transformations as well as the only t-distribution covariance for multi-view registration. Since only a few model parameters require to be optimized, the proposed method is more likely to obtain the desired registration results. Besides, all t-distribution centroids can be obtained by the NN search method, it is very efficient to achieve multi-view registration. What's more, the t-distribution takes the noise with heavy-tail into consideration, which makes the proposed method be inherently robust to noises and outliers. Experimental results tested on benchmark data sets illustrate its superior performance on robustness and accuracy over state-of-the-art methods.

59.Bi-Classifier Determinacy Maximization for Unsupervised Domain Adaptation ⬇️

Unsupervised domain adaptation challenges the problem of transferring knowledge from a well-labelled source domain to an unlabelled target domain. Recently,adversarial learning with bi-classifier has been proven effective in pushing cross-domain distributions close. Prior approaches typically leverage the disagreement between bi-classifier to learn transferable representations, however, they often neglect the classifier determinacy in the target domain, which could result in a lack of feature discriminability. In this paper, we present a simple yet effective method, namely Bi-Classifier Determinacy Maximization(BCDM), to tackle this problem. Motivated by the observation that target samples cannot always be separated distinctly by the decision boundary, here in the proposed BCDM, we design a novel classifier determinacy disparity (CDD) metric, which formulates classifier discrepancy as the class relevance of distinct target predictions and implicitly introduces constraint on the target feature discriminability. To this end, the BCDM can generate discriminative representations by encouraging target predictive outputs to be consistent and determined, meanwhile, preserve the diversity of predictions in an adversarial manner. Furthermore, the properties of CDD as well as the theoretical guarantees of BCDM's generalization bound are both elaborated. Extensive experiments show that BCDM compares favorably against the existing state-of-the-art domain adaptation methods.

60.Contrastive Learning for Label-Efficient Semantic Segmentation ⬇️

Collecting labeled data for the task of semantic segmentation is expensive and time-consuming, as it requires dense pixel-level annotations. While recent Convolutional Neural Network (CNN) based semantic segmentation approaches have achieved impressive results by using large amounts of labeled training data, their performance drops significantly as the amount of labeled data decreases. This happens because deep CNNs trained with the de facto cross-entropy loss can easily overfit to small amounts of labeled data. To address this issue, we propose a simple and effective contrastive learning-based training strategy in which we first pretrain the network using a pixel-wise class label-based contrastive loss, and then fine-tune it using the cross-entropy loss. This approach increases intra-class compactness and inter-class separability thereby resulting in a better pixel classifier. We demonstrate the effectiveness of the proposed training strategy in both fully-supervised and semi-supervised settings using the Cityscapes and PASCAL VOC 2012 segmentation datasets. Our results show that pretraining with label-based contrastive loss results in large performance gains (more than 20% absolute improvement in some settings) when the amount of labeled data is limited.

61.GeoNet++: Iterative Geometric Neural Network with Edge-Aware Refinement for Joint Depth and Surface Normal Estimation ⬇️

In this paper, we propose a geometric neural network with edge-aware refinement (GeoNet++) to jointly predict both depth and surface normal maps from a single image. Building on top of two-stream CNNs, GeoNet++ captures the geometric relationships between depth and surface normals with the proposed depth-to-normal and normal-to-depth modules. In particular, the "depth-to-normal" module exploits the least square solution of estimating surface normals from depth to improve their quality, while the "normal-to-depth" module refines the depth map based on the constraints on surface normals through kernel regression. Boundary information is exploited via an edge-aware refinement module. GeoNet++ effectively predicts depth and surface normals with strong 3D consistency and sharp boundaries resulting in better reconstructed 3D scenes. Note that GeoNet++ is generic and can be used in other depth/normal prediction frameworks to improve the quality of 3D reconstruction and pixel-wise accuracy of depth and surface normals. Furthermore, we propose a new 3D geometric metric (3DGM) for evaluating depth prediction in 3D. In contrast to current metrics that focus on evaluating pixel-wise error/accuracy, 3DGM measures whether the predicted depth can reconstruct high-quality 3D surface normals. This is a more natural metric for many 3D application domains. Our experiments on NYUD-V2 and KITTI datasets verify that GeoNet++ produces fine boundary details, and the predicted depth can be used to reconstruct high-quality 3D surfaces. Code has been made publicly available.

62.MVFNet: Multi-View Fusion Network for Efficient Video Recognition ⬇️

Conventionally, spatiotemporal modeling network and its complexity are the two most concentrated research topics in video action recognition. Existing state-of-the-art methods have achieved excellent accuracy regardless of the complexity meanwhile efficient spatiotemporal modeling solutions are slightly inferior in performance. In this paper, we attempt to acquire both efficiency and effectiveness simultaneously. First of all, besides traditionally treating H x W x T video frames as space-time signal (viewing from the Height-Width spatial plane), we propose to also model video from the other two Height-Time and Width-Time planes, to capture the dynamics of video thoroughly. Secondly, our model is designed based on 2D CNN backbones and model complexity is well kept in mind by design. Specifically, we introduce a novel multi-view fusion (MVF) module to exploit video dynamics using separable convolution for efficiency. It is a plug-and-play module and can be inserted into off-the-shelf 2D CNNs to form a simple yet effective model called MVFNet. Moreover, MVFNet can be thought of as a generalized video modeling framework and it can specialize to be existing methods such as C2D, SlowOnly, and TSM under different settings. Extensive experiments are conducted on popular benchmarks (i.e., Something-Something V1 & V2, Kinetics, UCF-101, and HMDB-51) to show its superiority. The proposed MVFNet can achieve state-of-the-art performance with 2D CNN's complexity.

63.Spontaneous Emotion Recognition from Facial Thermal Images ⬇️

One of the key research areas in computer vision addressed by a vast number of publications is the processing and understanding of images containing human faces. The most often addressed tasks include face detection, facial landmark localization, face recognition and facial expression analysis. Other, more specialized tasks such as affective computing, the extraction of vital signs from videos or analysis of social interaction usually require one or several of the aforementioned tasks that have to be performed. In our work, we analyze that a large number of tasks for facial image processing in thermal infrared images that are currently solved using specialized rule-based methods or not solved at all can be addressed with modern learning-based approaches. We have used USTC-NVIE database for training of a number of machine learning algorithms for facial landmark localization.

64.Fully-Automated Liver Tumor Localization and Characterization from Multi-Phase MR Volumes Using Key-Slice ROI Parsing: A Physician-Inspired Approach ⬇️

Using radiological scans to identify liver tumors is crucial for proper patient treatment. This is highly challenging, as top radiologists only achieve F1 scores of roughly 80% (hepatocellular carcinoma (HCC) vs. others) with only moderate inter-rater agreement, even when using multi-phase magnetic resonance (MR) imagery. Thus, there is great impetus for computer-aided diagnosis (CAD) solutions. A critical challengeis to reliably parse a 3D MR volume to localize diagnosable regions of interest (ROI). In this paper, we break down this problem using a key-slice parser (KSP), which emulates physician workflows by first identifying key slices and then localize their corresponding key ROIs. Because performance demands are so extreme, (not to miss any key ROI),our KSP integrates complementary modules--top-down classification-plus-detection (CPD) and bottom-up localization-by-over-segmentation(LBOS). The CPD uses a curve-parsing and detection confidence to re-weight classifier confidences. The LBOS uses over-segmentation to flag CPD failure cases and provides its own ROIs. For scalability, LBOS is only weakly trained on pseudo-masks using a new distance-aware Tversky loss. We evaluate our approach on the largest multi-phase MR liver lesion test dataset to date (430 biopsy-confirmed patients). Experiments demonstrate that our KSP can localize diagnosable ROIs with high reliability (85% patients have an average overlap of >= 40% with the ground truth). Moreover, we achieve an HCC vs. others F1 score of 0.804, providing a fully-automated CAD solution comparable with top human physicians.

65.Using Computer Vision to Automate Hand Detection and Tracking of Surgeon Movements in Videos of Open Surgery ⬇️

Open, or non-laparoscopic surgery, represents the vast majority of all operating room procedures, but few tools exist to objectively evaluate these techniques at scale. Current efforts involve human expert-based visual assessment. We leverage advances in computer vision to introduce an automated approach to video analysis of surgical execution. A state-of-the-art convolutional neural network architecture for object detection was used to detect operating hands in open surgery videos. Automated assessment was expanded by combining model predictions with a fast object tracker to enable surgeon-specific hand tracking. To train our model, we used publicly available videos of open surgery from YouTube and annotated these with spatial bounding boxes of operating hands. Our model's spatial detections of operating hands significantly outperforms the detections achieved using pre-existing hand-detection datasets, and allow for insights into intra-operative movement patterns and economy of motion.

66.MiniVLM: A Smaller and Faster Vision-Language Model ⬇️

Recent vision-language (VL) studies have shown remarkable progress by learning generic representations from massive image-text pairs with transformer models and then fine-tuning on downstream VL tasks. While existing research has been focused on achieving high accuracy with large pre-trained models, building a lightweight model is of great value in practice but is less explored. In this paper, we propose a smaller and faster VL model, MiniVLM, which can be finetuned with good performance on various downstream tasks like its larger counterpart. MiniVLM consists of two modules, a vision feature extractor and a transformer-based vision-language fusion module. We design a Two-stage Efficient feature Extractor (TEE), inspired by the one-stage EfficientDet network, to significantly reduce the time cost of visual feature extraction by $95%$, compared to a baseline model. We adopt the MiniLM structure to reduce the computation cost of the transformer module after comparing different compact BERT models. In addition, we improve the MiniVLM pre-training by adding $7M$ Open Images data, which are pseudo-labeled by a state-of-the-art captioning model. We also pre-train with high-quality image tags obtained from a strong tagging model to enhance cross-modality alignment. The large models are used offline without adding any overhead in fine-tuning and inference. With the above design choices, our MiniVLM reduces the model size by $73%$ and the inference time cost by $94%$ while being able to retain $94-97%$ of the accuracy on multiple VL tasks. We hope that MiniVLM helps ease the use of the state-of-the-art VL research for on-the-edge applications.

67.Human Pose Transfer by Adaptive Hierarchical Deformation ⬇️

Human pose transfer, as a misaligned image generation task, is very challenging. Existing methods cannot effectively utilize the input information, which often fail to preserve the style and shape of hair and clothes. In this paper, we propose an adaptive human pose transfer network with two hierarchical deformation levels. The first level generates human semantic parsing aligned with the target pose, and the second level generates the final textured person image in the target pose with the semantic guidance. To avoid the drawback of vanilla convolution that treats all the pixels as valid information, we use gated convolution in both two levels to dynamically select the important features and adaptively deform the image layer by layer. Our model has very few parameters and is fast to converge. Experimental results demonstrate that our model achieves better performance with more consistent hair, face and clothes with fewer parameters than state-of-the-art methods. Furthermore, our method can be applied to clothing texture transfer.

68.Assessing The Importance Of Colours For CNNs In Object Recognition ⬇️

Humans rely heavily on shapes as a primary cue for object recognition. As secondary cues, colours and textures are also beneficial in this regard. Convolutional neural networks (CNNs), an imitation of biological neural networks, have been shown to exhibit conflicting properties. Some studies indicate that CNNs are biased towards textures whereas, another set of studies suggests shape bias for a classification task. However, they do not discuss the role of colours, implying its possible humble role in the task of object recognition. In this paper, we empirically investigate the importance of colours in object recognition for CNNs. We are able to demonstrate that CNNs often rely heavily on colour information while making a prediction. Our results show that the degree of dependency on colours tend to vary from one dataset to another. Moreover, networks tend to rely more on colours if trained from scratch. Pre-training can allow the model to be less colour dependent. To facilitate these findings, we follow the framework often deployed in understanding role of colours in object recognition for humans. We evaluate a model trained with congruent images (images in original colours eg. red strawberries) on congruent, greyscale, and incongruent images (images in unnatural colours eg. blue strawberries). We measure and analyse network's predictive performance (top-1 accuracy) under these different stylisations. We utilise standard datasets of supervised image classification and fine-grained image classification in our experiments.

69.PAIRS AutoGeo: an Automated Machine Learning Framework for Massive Geospatial Data ⬇️

An automated machine learning framework for geospatial data named PAIRS AutoGeo is introduced on IBM PAIRS Geoscope big data and analytics platform. The framework simplifies the development of industrial machine learning solutions leveraging geospatial data to the extent that the user inputs are minimized to merely a text file containing labeled GPS coordinates. PAIRS AutoGeo automatically gathers required data at the location coordinates, assembles the training data, performs quality check, and trains multiple machine learning models for subsequent deployment. The framework is validated using a realistic industrial use case of tree species classification. Open-source tree species data are used as the input to train a random forest classifier and a modified ResNet model for 10-way tree species classification based on aerial imagery, which leads to an accuracy of $59.8%$ and $81.4%$, respectively. This use case exemplifies how PAIRS AutoGeo enables users to leverage machine learning without extensive geospatial expertise.

70.AMINN: Autoencoder-based Multiple Instance Neural Network for Outcome Prediction of Multifocal Liver Metastases ⬇️

Colorectal cancer is one of the most common and lethal cancers and colorectal cancer liver metastases (CRLM) is the major cause of death in patients with colorectal cancer. Multifocality occurs frequently in CRLM, but is relatively unexplored in CRLM outcome prediction. Most existing clinical and imaging biomarkers do not take the imaging features of all multifocal lesions into account. In this paper, we present an end-to-end autoencoder-based multiple instance neural network (AMINN) for the prediction of survival outcomes in multifocal CRLM patients using radiomic features extracted from contrast-enhanced MRIs. Specifically, we jointly train an autoencoder to reconstruct input features and a multiple instance network to make predictions by aggregating information from all tumour lesions of a patient. In addition, we incorporate a two-step normalization technique to improve the training of deep neural networks, built on the observation that the distributions of radiomic features are almost always severely skewed. Experimental results empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. The proposed ADMINN framework achieved an area under the ROC curve (AUC) of 0.70, which is 19.5% higher than baseline methods. We built a risk score based on the outputs of our network and compared it to other clinical and imaging biomarkers. Our risk score is the only one that achieved statistical significance in univariate and multivariate cox proportional hazard modeling in our cohort of multifocal CRLM patients. The effectiveness of incorporating all lesions and applying two-step normalization is demonstrated by a series of ablation studies. Our code will be released after the peer-review process.

71.Spectral Unmixing With Multinomial Mixture Kernel and Wasserstein Generative Adversarial Loss ⬇️

This study proposes a novel framework for spectral unmixing by using 1D convolution kernels and spectral uncertainty. High-level representations are computed from data, and they are further modeled with the Multinomial Mixture Model to estimate fractions under severe spectral uncertainty. Furthermore, a new trainable uncertainty term based on a nonlinear neural network model is introduced in the reconstruction step. All uncertainty models are optimized by Wasserstein Generative Adversarial Network (WGAN) to improve stability and capture uncertainty. Experiments are performed on both real and synthetic datasets. The results validate that the proposed method obtains state-of-the-art performance, especially for the real datasets compared to the baselines. Project page at: this https URL.

72.LiveChess2FEN: a Framework for Classifying Chess Pieces based on CNNs ⬇️

Automatic digitization of chess games using computer vision is a significant technological challenge. This problem is of much interest for tournament organizers and amateur or professional players to broadcast their over-the-board (OTB) games online or analyze them using chess engines. Previous work has shown promising results, but the recognition accuracy and the latency of state-of-the-art techniques still need further enhancements to allow their practical and affordable deployment. We have investigated how to implement them on an Nvidia Jetson Nano single-board computer effectively. Our first contribution has been accelerating the chessboard's detection algorithm. Subsequently, we have analyzed different Convolutional Neural Networks for chess piece classification and how to map them efficiently on our embedded platform. Notably, we have implemented a functional framework that automatically digitizes a chess position from an image in less than 1 second, with 92% accuracy when classifying the pieces and 95% when detecting the board.

73.Multi-Scale Cascading Network with Compact Feature Learning for RGB-Infrared Person Re-Identification ⬇️

RGB-Infrared person re-identification (RGB-IR Re-ID) aims to match persons from heterogeneous images captured by visible and thermal cameras, which is of great significance in the surveillance system under poor light conditions. Facing great challenges in complex variances including conventional single-modality and additional inter-modality discrepancies, most of the existing RGB-IR Re-ID methods propose to impose constraints in image level, feature level or a hybrid of both. Despite the better performance of hybrid constraints, they are usually implemented with heavy network architecture. As a matter of fact, previous efforts contribute more as pioneering works in new cross-modal Re-ID area while leaving large space for improvement. This can be mainly attributed to: (1) lack of abundant person image pairs from different modalities for training, and (2) scarcity of salient modality-invariant features especially on coarse representations for effective matching. To address these issues, a novel Multi-Scale Part-Aware Cascading framework (MSPAC) is formulated by aggregating multi-scale fine-grained features from part to global in a cascading manner, which results in a unified representation containing rich and enhanced semantic features. Furthermore, a marginal exponential centre (MeCen) loss is introduced to jointly eliminate mixed variances from intra- and inter-modal examples. Cross-modality correlations can thus be efficiently explored on salient features for distinctive modality-invariant feature learning. Extensive experiments are conducted to demonstrate that the proposed method outperforms all the state-of-the-art by a large margin.

74.High Order Local Directional Pattern Based Pyramidal Multi-structure for Robust Face Recognition ⬇️

Derived from a general definition of texture in a local neighborhood, local directional pattern (LDP) encodes the directional information in the small local 3x3 neighborhood of a pixel, which may fail to extract detailed information especially during changes in the input image due to illumination variations. Therefore, in this paper we introduce a novel feature extraction technique that calculates the nth order direction variation patterns, named high order local directional pattern (HOLDP). The proposed HOLDP can capture more detailed discriminative information than the conventional LDP. Unlike the LDP operator, our proposed technique extracts nth order local information by encoding various distinctive spatial relationships from each neighborhood layer of a pixel in the pyramidal multi-structure way. Then we concatenate the feature vector of each neighborhood layer to form the final HOLDP feature vector. The performance evaluation of the proposed HOLDP algorithm is conducted on several publicly available face databases and observed the superiority of HOLDP under extreme illumination conditions.

75.Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation ⬇️

Visual object tracking aims to precisely estimate the bounding box for the given target, which is a challenging problem due to factors such as deformation and occlusion. Many recent trackers adopt the multiple-stage tracking strategy to improve the quality of bounding box estimation. These methods first coarsely locate the target and then refine the initial prediction in the following stages. However, existing approaches still suffer from limited precision, and the coupling of different stages severely restricts the method's transferability. This work proposes a novel, flexible, and accurate refinement module called Alpha-Refine, which can significantly improve the base trackers' prediction quality. By exploring a series of design options, we conclude that the key to successful refinement is extracting and maintaining detailed spatial information as much as possible. Following this principle, Alpha-Refine adopts a pixel-wise correlation, a corner prediction head, and an auxiliary mask head as the core components. We apply Alpha-Refine to six famous base trackers to verify our method's effectiveness: DiMPsuper, DiMP50, ATOM, SiamRPN++, RT-MDNet, and ECO. Comprehensive experiments on TrackingNet, LaSOT, GOT-10K, and VOT2020 benchmarks show that our approach significantly improves the base tracker's performance with little extra latency. Code and pretrained model is available at this https URL.

76.Fine-grained Classification via Categorical Memory Networks ⬇️

Motivated by the desire to exploit patterns shared across classes, we present a simple yet effective class-specific memory module for fine-grained feature learning. The memory module stores the prototypical feature representation for each category as a moving average. We hypothesize that the combination of similarities with respect to each category is itself a useful discriminative cue. To detect these similarities, we use attention as a querying mechanism. The attention scores with respect to each class prototype are used as weights to combine prototypes via weighted sum, producing a uniquely tailored response feature representation for a given input. The original and response features are combined to produce an augmented feature for classification. We integrate our class-specific memory module into a standard convolutional neural network, yielding a Categorical Memory Network. Our memory module significantly improves accuracy over baseline CNNs, achieving competitive accuracy with state-of-the-art methods on four benchmarks, including CUB-200-2011, Stanford Cars, FGVC Aircraft, and NABirds.

77.DETR for Pedestrian Detection ⬇️

Pedestrian detection in crowd scenes poses a challenging problem due to the heuristic defined mapping from anchors to pedestrians and the conflict between NMS and highly overlapped pedestrians. The recently proposed end-to-end detectors(ED), DETR and deformable DETR, replace hand designed components such as NMS and anchors using the transformer architecture, which gets rid of duplicate predictions by computing all pairwise interactions between queries. Inspired by these works, we explore their performance on crowd pedestrian detection. Surprisingly, compared to Faster-RCNN with FPN, the results are opposite to those obtained on COCO. Furthermore, the bipartite match of ED harms the training efficiency due to the large ground truth number in crowd scenes. In this work, we identify the underlying motives driving ED's poor performance and propose a new decoder to address them. Moreover, we design a mechanism to leverage the less occluded visible parts of pedestrian specifically for ED, and achieve further improvements. A faster bipartite match algorithm is also introduced to make ED training on crowd dataset more practical. The proposed detector PED(Pedestrian End-to-end Detector) outperforms both previous EDs and the baseline Faster-RCNN on CityPersons and CrowdHuman. It also achieves comparable performance with state-of-the-art pedestrian detection methods. Code will be released soon.

78.Uncalibrated Neural Inverse Rendering for Photometric Stereo of General Surfaces ⬇️

This paper presents an uncalibrated deep neural network framework for the photometric stereo problem. For training models to solve the problem, existing neural network-based methods either require exact light directions or ground-truth surface normals of the object or both. However, in practice, it is challenging to procure both of this information precisely, which restricts the broader adoption of photometric stereo algorithms for vision application. To bypass this difficulty, we propose an uncalibrated neural inverse rendering approach to this problem. Our method first estimates the light directions from the input images and then optimizes an image reconstruction loss to calculate the surface normals, bidirectional reflectance distribution function value, and depth. Additionally, our formulation explicitly models the concave and convex parts of a complex surface to consider the effects of interreflections in the image formation process. Extensive evaluation of the proposed method on the challenging subjects generally shows comparable or better results than the supervised and classical approaches.

79.An Overview of Depth Cameras and Range Scanners Based on Time-of-Flight Technologies ⬇️

Time-of-flight (TOF) cameras are sensors that can measure the depths of scene-points, by illuminating the scene with a controlled laser or LED source, and then analyzing the reflected light. In this paper, we will first describe the underlying measurement principles of time-of-flight cameras, including: (i) pulsed-light cameras, which measure directly the time taken for a light pulse to travel from the device to the object and back again, and (ii) continuous-wave modulated-light cameras, which measure the phase difference between the emitted and received signals, and hence obtain the travel time indirectly. We review the main existing designs, including prototypes as well as commercially available devices. We also review the relevant camera calibration principles, and how they are applied to TOF devices. Finally, we discuss the benefits and challenges of combined TOF and color camera systems.

80.Fusion of Range and Stereo Data for High-Resolution Scene-Modeling ⬇️

This paper addresses the problem of range-stereo fusion, for the construction of high-resolution depth maps. In particular, we combine low-resolution depth data with high-resolution stereo data, in a maximum a posteriori (MAP) formulation. Unlike existing schemes that build on MRF optimizers, we infer the disparity map from a series of local energy minimization problems that are solved hierarchically, by growing sparse initial disparities obtained from the depth data. The accuracy of the method is not compromised, owing to three properties of the data-term in the energy function. Firstly, it incorporates a new correlation function that is capable of providing refined correlations and disparities, via subpixel correction. Secondly, the correlation scores rely on an adaptive cost aggregation step, based on the depth data. Thirdly, the stereo and depth likelihoods are adaptively fused, based on the scene texture and camera geometry. These properties lead to a more selective growing process which, unlike previous seed-growing methods, avoids the tendency to propagate incorrect disparities. The proposed method gives rise to an intrinsically efficient algorithm, which runs at 3FPS on 2.0MP images on a standard desktop computer. The strong performance of the new method is established both by quantitative comparisons with state-of-the-art methods, and by qualitative comparisons using real depth-stereo data-sets.

81.Anomaly detection through latent space restoration using vector-quantized variational autoencoders ⬇️

We propose an out-of-distribution detection method that combines density and restoration-based approaches using Vector-Quantized Variational Auto-Encoders (VQ-VAEs). The VQ-VAE model learns to encode images in a categorical latent space. The prior distribution of latent codes is then modelled using an Auto-Regressive (AR) model. We found that the prior probability estimated by the AR model can be useful for unsupervised anomaly detection and enables the estimation of both sample and pixel-wise anomaly scores. The sample-wise score is defined as the negative log-likelihood of the latent variables above a threshold selecting highly unlikely codes. Additionally, out-of-distribution images are restored into in-distribution images by replacing unlikely latent codes with samples from the prior model and decoding to pixel space. The average L1 distance between generated restorations and original image is used as pixel-wise anomaly score. We tested our approach on the MOOD challenge datasets, and report higher accuracies compared to a standard reconstruction-based approach with VAEs.

82.Periocular in the Wild Embedding Learning with Cross-Modal Consistent Knowledge Distillation ⬇️

Periocular biometric, or peripheral area of ocular, is a collaborative alternative to face, especially if a face is occluded or masked. In practice, sole periocular biometric captures least salient facial features, thereby suffering from intra-class compactness and inter-class dispersion issues particularly in the wild environment. To address these problems, we transfer useful information from face to support periocular modality by means of knowledge distillation (KD) for embedding learning. However, applying typical KD techniques to heterogeneous modalities directly is suboptimal. We put forward in this paper a deep face-to-periocular distillation networks, coined as cross-modal consistent knowledge distillation (CM-CKD) henceforward. The three key ingredients of CM-CKD are (1) shared-weight networks, (2) consistent batch normalization, and (3) a bidirectional consistency distillation for face and periocular through an effectual CKD loss. To be more specific, we leverage face modality for periocular embedding learning, but only periocular images are targeted for identification or verification tasks. Extensive experiments on six constrained and unconstrained periocular datasets disclose that the CM-CKD-learned periocular embeddings extend identification and verification performance by 50% in terms of relative performance gain computed based upon face and periocular baselines. The experiments also reveal that the CM-CKD-learned periocular features enjoy better subject-wise cluster separation, thereby refining the overall accuracy performance.

83.Computer Vision and Normalizing Flow Based Defect Detection ⬇️

Surface defect detection is essential and necessary for controlling the qualities of the products during manufacturing. The challenges in this complex task include: 1) collecting defective samples and manually labeling for training is time-consuming; 2) the defects' characteristics are difficult to define as new types of defect can happen all the time; 3) and the real-world product images contain lots of background noise. In this paper, we present a two-stage defect detection network based on the object detection model YOLO, and the normalizing flow-based defect detection model DifferNet. Our model has high robustness and performance on defect detection using real-world video clips taken from a production line monitoring system. The normalizing flow-based anomaly detection model only requires a small number of good samples for training and then perform defect detection on the product images detected by YOLO. The model we invent employs two novel strategies: 1) a two-stage network using YOLO and a normalizing flow-based model to perform product defect detection, 2) multi-scale image transformations are implemented to solve the issue product image cropped by YOLO includes many background noise. Besides, extensive experiments are conducted on a new dataset collected from the real-world factory production line. We demonstrate that our proposed model can learn on a small number of defect-free samples of single or multiple product types. The dataset will also be made public to encourage further studies and research in surface defect detection.

84.Multimodal In-bed Pose and Shape Estimation under the Blankets ⬇️

Humans spend vast hours in bed -- about one-third of the lifetime on average. Besides, a human at rest is vital in many healthcare applications. Typically, humans are covered by a blanket when resting, for which we propose a multimodal approach to uncover the subjects so their bodies at rest can be viewed without the occlusion of the blankets above. We propose a pyramid scheme to effectively fuse the different modalities in a way that best leverages the knowledge captured by the multimodal sensors. Specifically, the two most informative modalities (i.e., depth and infrared images) are first fused to generate good initial pose and shape estimation. Then pressure map and RGB images are further fused one by one to refine the result by providing occlusion-invariant information for the covered part, and accurate shape information for the uncovered part, respectively. However, even with multimodal data, the task of detecting human bodies at rest is still very challenging due to the extreme occlusion of bodies. To further reduce the negative effects of the occlusion from blankets, we employ an attention-based reconstruction module to generate uncovered modalities, which are further fused to update current estimation via a cyclic fashion. Extensive experiments validate the superiority of the proposed model over others.

85.PoP-Net: Pose over Parts Network for Multi-Person 3D Pose Estimation from a Depth Image ⬇️

In this paper, a real-time method called PoP-Net is proposed to predict multi-person 3D poses from a depth image. PoP-Net learns to predict bottom-up part detection maps and top-down global poses in a single-shot framework. A simple and effective fusion process is applied to fuse the global poses and part detection. Specifically, a new part-level representation, called Truncated Part Displacement Field (TPDF), is introduced. It drags low-precision global poses towards more accurate part locations while maintaining the advantage of global poses in handling severe occlusion and truncation cases. A mode selection scheme is developed to automatically resolve the conflict between global poses and local detection. Finally, due to the lack of high-quality depth datasets for developing and evaluating multi-person 3D pose estimation methods, a comprehensive depth dataset with 3D pose labels is released. The dataset is designed to enable effective multi-person and background data augmentation such that the developed models are more generalizable towards uncontrolled real-world multi-person scenarios. We show that PoP-Net has significant advantages in efficiency for multi-person processing and achieves the state-of-the-art results both on the released challenging dataset and on the widely used ITOP dataset.

86.Mask Guided Matting via Progressive Refinement Network ⬇️

We propose Mask Guided (MG) Matting, a robust matting framework that takes a general coarse mask as guidance. MG Matting leverages a network (PRN) design which encourages the matting model to provide self-guidance to progressively refine the uncertain regions through the decoding process. A series of guidance mask perturbation operations are also introduced in the training to further enhance its robustness to external guidance. We show that PRN can generalize to unseen types of guidance masks such as trimap and low-quality alpha matte, making it suitable for various application pipelines. In addition, we revisit the foreground color prediction problem for matting and propose a surprisingly simple improvement to address the dataset issue. Evaluation on real and synthetic benchmarks shows that MG Matting achieves state-of-the-art performance using various types of guidance inputs. Code and models will be available at this https URL

87.Teacher-Student Asynchronous Learning with Multi-Source Consistency for Facial Landmark Detection ⬇️

Due to the high annotation cost of large-scale facial landmark detection tasks in videos, a semi-supervised paradigm that uses self-training for mining high-quality pseudo-labels to participate in training has been proposed by researchers. However, self-training based methods often train with a gradually increasing number of samples, whose performances vary a lot depending on the number of pseudo-labeled samples added.
In this paper, we propose a teacher-student asynchronous learning~(TSAL) framework based on the multi-source supervision signal consistency criterion, which implicitly mines pseudo-labels through consistency constraints. Specifically, the TSAL framework contains two models with exactly the same structure. The radical student uses multi-source supervision signals from the same task to update parameters, while the calm teacher uses a single-source supervision signal to update parameters. In order to reasonably absorb student's suggestions, teacher's parameters are updated again through recursive average filtering. The experimental results prove that asynchronous-learning framework can effectively filter noise in multi-source supervision signals, thereby mining the pseudo-labels which are more significant for network parameter updating. And extensive experiments on 300W, AFLW, and 300VW benchmarks show that the TSAL framework achieves state-of-the-art performance.

88.D$^2$IM-Net: Learning Detail Disentangled Implicit Fields from Single Images ⬇️

We present the first single-view 3D reconstruction network aimed at recovering geometric details from an input image which encompass both topological shape structures and surface features. Our key idea is to train the network to learn a detail disentangled reconstruction consisting of two functions, one implicit field representing the coarse 3D shape and the other capturing the details. Given an input image, our network, coined D$^2$IM-Net, encodes it into global and local features which are respectively fed into two decoders. The base decoder uses the global features to reconstruct a coarse implicit field, while the detail decoder reconstructs, from the local features, two displacement maps, defined over the front and back sides of the captured object. The final 3D reconstruction is a fusion between the base shape and the displacement maps, with three losses enforcing the recovery of coarse shape, overall structure, and surface details via a novel Laplacian term.

89.Street-view Panoramic Video Synthesis from a Single Satellite Image ⬇️

We present a novel method for synthesizing both temporally and geometrically consistent street-view panoramic video from a given single satellite image and camera trajectory. Existing cross-view synthesis approaches focus more on images, while video synthesis in such a case has not yet received enough attention. Single image synthesis approaches are not well suited for video synthesis since they lack temporal consistency which is a crucial property of videos. To this end, our approach explicitly creates a 3D point cloud representation of the scene and maintains dense 3D-2D correspondences across frames that reflect the geometric scene configuration inferred from the satellite view. We implement a cascaded network architecture with two hourglass modules for successive coarse and fine generation for colorizing the point cloud from the semantics and per-class latent vectors. By leveraging computed correspondences, the produced street-view video frames adhere to the 3D geometric scene structure and maintain temporal consistency. Qualitative and quantitative experiments demonstrate superior results compared to other state-of-the-art cross-view synthesis approaches that either lack temporal or geometric consistency. To the best of our knowledge, our work is the first work to synthesize cross-view images to video.

90.Vilio: State-of-the-art Visio-Linguistic Models applied to Hateful Memes ⬇️

This work presents Vilio, an implementation of state-of-the-art visio-linguistic models and their application to the Hateful Memes Dataset. The implemented models have been fitted into a uniform code-base and altered to yield better performance. The goal of Vilio is to provide a user-friendly starting point for any visio-linguistic problem. An ensemble of 5 different V+L models implemented in Vilio achieves 2nd place in the Hateful Memes Challenge out of 3,300 participants. The code is available at this https URL.

91.Improving Adversarial Robustness via Probabilistically Compact Loss with Logit Constraints ⬇️

Convolutional neural networks (CNNs) have achieved state-of-the-art performance on various tasks in computer vision. However, recent studies demonstrate that these models are vulnerable to carefully crafted adversarial samples and suffer from a significant performance drop when predicting them. Many methods have been proposed to improve adversarial robustness (e.g., adversarial training and new loss functions to learn adversarially robust feature representations). Here we offer a unique insight into the predictive behavior of CNNs that they tend to misclassify adversarial samples into the most probable false classes. This inspires us to propose a new Probabilistically Compact (PC) loss with logit constraints which can be used as a drop-in replacement for cross-entropy (CE) loss to improve CNN's adversarial robustness. Specifically, PC loss enlarges the probability gaps between true class and false classes meanwhile the logit constraints prevent the gaps from being melted by a small perturbation. We extensively compare our method with the state-of-the-art using large scale datasets under both white-box and black-box attacks to demonstrate its effectiveness. The source codes are available from the following url: this https URL.

92.Biomechanical modelling of brain atrophy through deep learning ⬇️

We present a proof-of-concept, deep learning (DL) based, differentiable biomechanical model of realistic brain deformations. Using prescribed maps of local atrophy and growth as input, the network learns to deform images according to a Neo-Hookean model of tissue deformation. The tool is validated using longitudinal brain atrophy data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, and we demonstrate that the trained model is capable of rapidly simulating new brain deformations with minimal residuals. This method has the potential to be used in data augmentation or for the exploration of different causal hypotheses reflecting brain growth and atrophy.

93.Movie Summarization via Sparse Graph Construction ⬇️

We summarize full-length movies by creating shorter videos containing their most informative scenes. We explore the hypothesis that a summary can be created by assembling scenes which are turning points (TPs), i.e., key events in a movie that describe its storyline. We propose a model that identifies TP scenes by building a sparse movie graph that represents relations between scenes and is constructed using multimodal information. According to human judges, the summaries created by our approach are more informative and complete, and receive higher ratings, than the outputs of sequence-based models and general-purpose summarization algorithms. The induced graphs are interpretable, displaying different topology for different movie genres.

94.Phase Retrieval with Holography and Untrained Priors: Tackling the Challenges of Low-Photon Nanoscale Imaging ⬇️

Phase retrieval is the inverse problem of recovering a signal from magnitude-only Fourier measurements, and underlies numerous imaging modalities, such as Coherent Diffraction Imaging (CDI). A variant of this setup, known as holography, includes a reference object that is placed adjacent to the specimen of interest before measurements are collected. The resulting inverse problem, known as holographic phase retrieval, is well-known to have improved problem conditioning relative to the original. This innovation, i.e. Holographic CDI, becomes crucial at the nanoscale, where imaging specimens such as viruses, proteins, and crystals require low-photon measurements. This data is highly corrupted by Poisson shot noise, and often lacks low-frequency content as well. In this work, we introduce a dataset-free deep learning framework for holographic phase retrieval adapted to these challenges. The key ingredients of our approach are the explicit and flexible incorporation of the physical forward model into an automatic differentiation procedure, the Poisson log-likelihood objective function, and an optional untrained deep image prior. We perform extensive evaluation under realistic conditions. Compared to competing classical methods, our method recovers signal from higher noise levels and is more resilient to suboptimal reference design, as well as to large missing regions of low frequencies in the observations. To the best of our knowledge, this is the first work to consider a dataset-free machine learning approach for holographic phase retrieval.

95.Learning Hybrid Representations for Automatic 3D Vessel Centerline Extraction ⬇️

Automatic blood vessel extraction from 3D medical images is crucial for vascular disease diagnoses. Existing methods based on convolutional neural networks (CNNs) may suffer from discontinuities of extracted vessels when segmenting such thin tubular structures from 3D images. We argue that preserving the continuity of extracted vessels requires to take into account the global geometry. However, 3D convolutions are computationally inefficient, which prohibits the 3D CNNs from sufficiently large receptive fields to capture the global cues in the entire image. In this work, we propose a hybrid representation learning approach to address this challenge. The main idea is to use CNNs to learn local appearances of vessels in image crops while using another point-cloud network to learn the global geometry of vessels in the entire image. In inference, the proposed approach extracts local segments of vessels using CNNs, classifies each segment based on global geometry using the point-cloud network, and finally connects all the segments that belong to the same vessel using the shortest-path algorithm. This combination results in an efficient, fully-automatic and template-free approach to centerline extraction from 3D images. We validate the proposed approach on CTA datasets and demonstrate its superior performance compared to both traditional and CNN-based baselines.

96.IPN-V2 and OCTA-500: Methodology and Dataset for Retinal Image Segmentation ⬇️

Optical coherence tomography angiography (OCTA) is a novel imaging modality that allows a micron-level resolution to present the three-dimensional structure of the retinal vascular. In our previous work, a 3D-to-2D image projection network (IPN) was proposed for retinal vessel (RV) and foveal avascular zone (FAZ) segmentations in OCTA images. One of its advantages is that the segmentation results are directly from the original volumes without using any projection images and retinal layer segmentation. In this work, we propose image projection network V2 (IPN-V2), extending IPN by adding a plane perceptron to enhance the perceptron ability in the horizontal direction. We also propose IPN-V2+, as a supplement of the IPN-V2, by introducing a global retraining process to overcome the "checkerboard effect". Besides, we propose a new multi-modality dataset, dubbed OCTA-500. It contains 500 subjects with two field of view (FOV) types, including OCT and OCTA volumes, six types of projections, four types of text labels and two types of pixel-level labels. The dataset contains more than 360K images with a size of about 80GB. To the best of our knowledge, it is currently the largest OCTA dataset with the abundant information. Finally, we perform a thorough evaluation of the performance of IPN-V2 on the OCTA-500 dataset. The experimental results demonstrate that our proposed IPN-V2 performs better than IPN and other deep learning methods in RV segmentation and FAZ segmentation.

97.Accurate Cell Segmentation in Digital Pathology Images via Attention Enforced Networks ⬇️

Automatic cell segmentation is an essential step in the pipeline of computer-aided diagnosis (CAD), such as the detection and grading of breast cancer. Accurate segmentation of cells can not only assist the pathologists to make a more precise diagnosis, but also save much time and labor. However, this task suffers from stain variation, cell inhomogeneous intensities, background clutters and cells from different tissues. To address these issues, we propose an Attention Enforced Network (AENet), which is built on spatial attention module and channel attention module, to integrate local features with global dependencies and weight effective channels adaptively. Besides, we introduce a feature fusion branch to bridge high-level and low-level features. Finally, the marker controlled watershed algorithm is applied to post-process the predicted segmentation maps for reducing the fragmented regions. In the test stage, we present an individual color normalization method to deal with the stain variation problem. We evaluate this model on the MoNuSeg dataset. The quantitative comparisons against several prior methods demonstrate the superiority of our approach.

98.Multi-Domain Multi-Task Rehearsal for Lifelong Learning ⬇️

Rehearsal, seeking to remind the model by storing old knowledge in lifelong learning, is one of the most effective ways to mitigate catastrophic forgetting, i.e., biased forgetting of previous knowledge when moving to new tasks. However, the old tasks of the most previous rehearsal-based methods suffer from the unpredictable domain shift when training the new task. This is because these methods always ignore two significant factors. First, the Data Imbalance between the new task and old tasks that makes the domain of old tasks prone to shift. Second, the Task Isolation among all tasks will make the domain shift toward unpredictable directions; To address the unpredictable domain shift, in this paper, we propose Multi-Domain Multi-Task (MDMT) rehearsal to train the old tasks and new task parallelly and equally to break the isolation among tasks. Specifically, a two-level angular margin loss is proposed to encourage the intra-class/task compactness and inter-class/task discrepancy, which keeps the model from domain chaos. In addition, to further address domain shift of the old tasks, we propose an optional episodic distillation loss on the memory to anchor the knowledge for each old task. Experiments on benchmark datasets validate the proposed approach can effectively mitigate the unpredictable domain shift.

99.D-LEMA: Deep Learning Ensembles from Multiple Annotations -- Application to Skin Lesion Segmentation ⬇️

Medical image segmentation annotations suffer from inter/intra-observer variations even among experts due to intrinsic differences in human annotators and ambiguous boundaries. Leveraging a collection of annotators' opinions for an image is an interesting way of estimating a gold standard. Although training deep models in a supervised setting with a single annotation per image has been extensively studied, generalizing their training to work with data sets containing multiple annotations per image remains a fairly unexplored problem. In this paper, we propose an approach to handle annotators' disagreements when training a deep model. To this end, we propose an ensemble of Bayesian fully convolutional networks (FCNs) for the segmentation task by considering two major factors in the aggregation of multiple ground truth annotations: (1) handling contradictory annotations in the training data originating from inter-annotator disagreements and (2) improving confidence calibration through the fusion of base models predictions. We demonstrate the superior performance of our approach on the ISIC Archive and explore the generalization performance of our proposed method by cross-data set evaluation on the PH2 and DermoFit data sets.

100.Pseudo Shots: Few-Shot Learning with Auxiliary Data ⬇️

In many practical few-shot learning problems, even though labeled examples are scarce, there are abundant auxiliary data sets that potentially contain useful information. We propose a framework to address the challenges of efficiently selecting and effectively using auxiliary data in image classification. Given an auxiliary dataset and a notion of semantic similarity among classes, we automatically select pseudo shots, which are labeled examples from other classes related to the target task. We show that naively assuming that these additional examples come from the same distribution as the target task examples does not significantly improve accuracy. Instead, we propose a masking module that adjusts the features of auxiliary data to be more similar to those of the target classes. We show that this masking module can improve accuracy by up to 18 accuracy points, particularly when the auxiliary data is semantically distant from the target task. We also show that incorporating pseudo shots improves over the current state-of-the-art few-shot image classification scores by an average of 4.81 percentage points of accuracy on 1-shot tasks and an average of 0.31 percentage points on 5-shot tasks.

101.Learning Contextual Causality from Time-consecutive Images ⬇️

Causality knowledge is crucial for many artificial intelligence systems. Conventional textual-based causality knowledge acquisition methods typically require laborious and expensive human annotations. As a result, their scale is often limited. Moreover, as no context is provided during the annotation, the resulting causality knowledge records (e.g., ConceptNet) typically do not take the context into consideration. To explore a more scalable way of acquiring causality knowledge, in this paper, we jump out of the textual domain and investigate the possibility of learning contextual causality from the visual signal. Compared with pure text-based approaches, learning causality from the visual signal has the following advantages: (1) Causality knowledge belongs to the commonsense knowledge, which is rarely expressed in the text but rich in videos; (2) Most events in the video are naturally time-ordered, which provides a rich resource for us to mine causality knowledge from; (3) All the objects in the video can be used as context to study the contextual property of causal relations. In detail, we first propose a high-quality dataset Vis-Causal and then conduct experiments to demonstrate that with good language and visual representation models as well as enough training signals, it is possible to automatically discover meaningful causal knowledge from the videos. Further analysis also shows that the contextual property of causal relations indeed exists, taking which into consideration might be crucial if we want to use the causality knowledge in real applications, and the visual signal could serve as a good resource for learning such contextual causality.

102.Robust Segmentation of Optic Disc and Cup from Fundus Images Using Deep Neural Networks ⬇️

Optic disc (OD) and optic cup (OC) are regions of prominent clinical interest in a retinal fundus image. They are the primary indicators of a glaucomatous condition. With the advent and success of deep learning for healthcare research, several approaches have been proposed for the segmentation of important features in retinal fundus images. We propose a novel approach for the simultaneous segmentation of the OD and OC using a residual encoder-decoder network (REDNet) based regional convolutional neural network (RCNN). The RED-RCNN is motivated by the Mask RCNN (MRCNN). Performance comparisons with the state-of-the-art techniques and extensive validations on standard publicly available fundus image datasets show that RED-RCNN has superior performance compared with MRCNN. RED-RCNN results in Sensitivity, Specificity, Accuracy, Precision, Dice and Jaccard indices of 95.64%, 99.9%, 99.82%, 95.68%, 95.64%, 91.65%, respectively, for OD segmentation, and 91.44%, 99.87%, 99.83%, 85.67%, 87.48%, 78.09%, respectively, for OC segmentation. Further, we perform two-stage glaucoma severity grading using the cup-to-disc ratio (CDR) computed based on the obtained OD/OC segmentation. The superior segmentation performance of RED-RCNN over MRCNN translates to higher accuracy in glaucoma severity grading.

103.Leaking Sensitive Financial Accounting Data in Plain Sight using Deep Autoencoder Neural Networks ⬇️

Nowadays, organizations collect vast quantities of sensitive information in Enterprise Resource Planning' (ERP) systems, such as accounting relevant transactions, customer master data, or strategic sales price information. The leakage of such information poses a severe threat for companies as the number of incidents and the reputational damage to those experiencing them continue to increase. At the same time, discoveries in deep learning research revealed that machine learning models could be maliciously misused to create new attack vectors. Understanding the nature of such attacks becomes increasingly important for the (internal) audit and fraud examination practice. The creation of such an awareness holds in particular for the fraudulent data leakage using deep learning-based steganographic techniques that might remain undetected by state-of-the-art Computer Assisted Audit Techniques' (CAATs). In this work, we introduce a real-world threat model' designed to leak sensitive accounting data. In addition, we show that a deep steganographic process, constituted by three neural networks, can be trained to hide such data in unobtrusive day-to-day' images. Finally, we provide qualitative and quantitative evaluations on two publicly available real-world payment datasets.

104.CHS-Net: A Deep learning approach for hierarchical segmentation of COVID-19 infected CT images ⬇️

The pandemic of novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) also known as COVID-19 has been spreading worldwide, causing rampant loss of lives. Medical imaging such as computed tomography (CT), X-ray, etc., plays a significant role in diagnosing the patients by presenting the visual representation of the functioning of the organs. However, for any radiologist analyzing such scans is a tedious and time-consuming task. The emerging deep learning technologies have displayed its strength in analyzing such scans to aid in the faster diagnosis of the diseases and viruses such as COVID-19. In the present article, an automated deep learning based model, COVID-19 hierarchical segmentation network (CHS-Net) is proposed that functions as a semantic hierarchical segmenter to identify the COVID-19 infected regions from lungs contour via CT medical imaging. The CHS-Net is developed with the two cascaded residual attention inception U-Net (RAIU-Net) models where first generates lungs contour maps and second generates COVID-19 infected regions. RAIU-Net comprises of a residual inception U-Net model with spectral spatial and depth attention network (SSD), consisting of contraction and expansion phases of depthwise separable convolutions and hybrid pooling (max and spectral pooling) to efficiently encode and decode the semantic and varying resolution information. The CHS-Net is trained with the segmentation loss function that is the weighted average of binary cross entropy loss and dice loss to penalize false negative and false positive predictions. The approach is compared with the recently proposed research works on the basis of standard metrics, it is observed that the proposed approach outperformed the recently proposed approaches and effectively segments the COVID-19 infected regions in the lungs.

105.LEARN++: Recurrent Dual-Domain Reconstruction Network for Compressed Sensing CT ⬇️

Compressed sensing (CS) computed tomography has been proven to be important for several clinical applications, such as sparse-view computed tomography (CT), digital tomosynthesis and interior tomography. Traditional compressed sensing focuses on the design of handcrafted prior regularizers, which are usually image-dependent and time-consuming. Inspired by recently proposed deep learning-based CT reconstruction models, we extend the state-of-the-art LEARN model to a dual-domain version, dubbed LEARN++. Different from existing iteration unrolling methods, which only involve projection data in the data consistency layer, the proposed LEARN++ model integrates two parallel and interactive subnetworks to perform image restoration and sinogram inpainting operations on both the image and projection domains simultaneously, which can fully explore the latent relations between projection data and reconstructed images. The experimental results demonstrate that the proposed LEARN++ model achieves competitive qualitative and quantitative results compared to several state-of-the-art methods in terms of both artifact reduction and detail preservation.

106.Learn-Prune-Share for Lifelong Learning ⬇️

In lifelong learning, we wish to maintain and update a model (e.g., a neural network classifier) in the presence of new classification tasks that arrive sequentially. In this paper, we propose a learn-prune-share (LPS) algorithm which addresses the challenges of catastrophic forgetting, parsimony, and knowledge reuse simultaneously. LPS splits the network into task-specific partitions via an ADMM-based pruning strategy. This leads to no forgetting, while maintaining parsimony. Moreover, LPS integrates a novel selective knowledge sharing scheme into this ADMM optimization framework. This enables adaptive knowledge sharing in an end-to-end fashion. Comprehensive experimental results on two lifelong learning benchmark datasets and a challenging real-world radio frequency fingerprinting dataset are provided to demonstrate the effectiveness of our approach. Our experiments show that LPS consistently outperforms multiple state-of-the-art competitors.

107.Attentional Biased Stochastic Gradient for Imbalanced Classification ⬇️

In this paper~\footnote{The original title is "Momentum SGD with Robust Weighting For Imbalanced Classification"}, we present a simple yet effective method (ABSGD) for addressing the data imbalance issue in deep learning. Our method is a simple modification to momentum SGD where we leverage an attentional mechanism to assign an individual importance weight to each gradient in the mini-batch. Unlike existing individual weighting methods that learn the individual weights by meta-learning on a separate balanced validation data, our weighting scheme is self-adaptive and is grounded in distributionally robust optimization. The weight of a sampled data is systematically proportional to exponential of a scaled loss value of the data, where the scaling factor is interpreted as the regularization parameter in the framework of information-regularized distributionally robust optimization. We employ a step damping strategy for the scaling factor to balance between the learning of feature extraction layers and the learning of the classifier layer. Compared with exiting meta-learning methods that require three backward propagations for computing mini-batch stochastic gradients at three different points at each iteration, our method is more efficient with only one backward propagation at each iteration as in standard deep learning methods. Compared with existing class-level weighting schemes, our method can be applied to online learning without any knowledge of class prior, while enjoying further performance boost in offline learning combined with existing class-level weighting schemes. Our empirical studies on several benchmark datasets also demonstrate the effectiveness of our proposed method

108.The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models ⬇️

The computer vision world has been re-gaining enthusiasm in various pre-trained models, including both classical ImageNet supervised pre-training and recently emerged self-supervised pre-training such as simCLR and MoCo. Pre-trained weights often boost a wide range of downstream tasks including classification, detection, and segmentation. Latest studies suggest that the pre-training benefits from gigantic model capacity. We are hereby curious and ask: after pre-training, does a pre-trained model indeed have to stay large for its universal downstream transferability?
In this paper, we examine the supervised and self-supervised pre-trained models through the lens of lottery ticket hypothesis (LTH). LTH identifies highly sparse matching subnetworks that can be trained in isolation from (nearly) scratch, to reach the full models' performance. We extend the scope of LTH to questioning whether matching subnetworks still exist in the pre-training models, that enjoy the same downstream transfer performance. Our extensive experiments convey an overall positive message: from all pre-trained weights obtained by ImageNet classification, simCLR and MoCo, we are consistently able to locate such matching subnetworks at 59.04% to 96.48% sparsity that transfer universally to multiple downstream tasks, whose performance see no degradation compared to using full pre-trained weights. Further analyses reveal that subnetworks found from different pre-training tend to yield diverse mask structures and perturbation sensitivities. We conclude that the core LTH observations remain generally relevant in the pre-training paradigm of computer vision, but more delicate discussions are needed in some cases. Codes and pre-trained models will be made available at: this https URL.

109.Interactive Radiotherapy Target Delineation with 3D-Fused Context Propagation ⬇️

Gross tumor volume (GTV) delineation on tomography medical imaging is crucial for radiotherapy planning and cancer diagnosis. Convolutional neural networks (CNNs) has been predominated on automatic 3D medical segmentation tasks, including contouring the radiotherapy target given 3D CT volume. While CNNs may provide feasible outcome, in clinical scenario, double-check and prediction refinement by experts is still necessary because of CNNs' inconsistent performance on unexpected patient cases. To provide experts an efficient way to modify the CNN predictions without retrain the model, we propose 3D-fused context propagation, which propagates any edited slice to the whole 3D volume. By considering the high-level feature maps, the radiation oncologists would only required to edit few slices to guide the correction and refine the whole prediction volume. Specifically, we leverage the backpropagation for activation technique to convey the user editing information backwardly to the latent space and generate new prediction based on the updated and original feature. During the interaction, our proposed approach reuses the extant extracted features and does not alter the existing 3D CNN model architectures, avoiding the perturbation on other predictions. The proposed method is evaluated on two published radiotherapy target contouring datasets of nasopharyngeal and esophageal cancer. The experimental results demonstrate that our proposed method is able to further effectively improve the existing segmentation prediction from different model architectures given oncologists' interactive inputs.

110.Delay Differential Neural Networks ⬇️

Neural ordinary differential equations (NODEs) treat computation of intermediate feature vectors as trajectories of ordinary differential equation parameterized by a neural network. In this paper, we propose a novel model, delay differential neural networks (DDNN), inspired by delay differential equations (DDEs). The proposed model considers the derivative of the hidden feature vector as a function of the current feature vector and past feature vectors (history). The function is modelled as a neural network and consequently, it leads to continuous depth alternatives to many recent ResNet variants. We propose two different DDNN architectures, depending on the way current and past feature vectors are considered. For training DDNNs, we provide a memory-efficient adjoint method for computing gradients and back-propagate through the network. DDNN improves the data efficiency of NODE by further reducing the number of parameters without affecting the generalization performance. Experiments conducted on synthetic and real-world image classification datasets such as Cifar10 and Cifar100 show the effectiveness of the proposed models.

111.Knowledge Capture and Replay for Continual Learning ⬇️

Deep neural networks have shown promise in several domains, and the learned task-specific information is implicitly stored in the network parameters. It will be vital to utilize representations from these networks for downstream tasks such as continual learning. In this paper, we introduce the notion of {\em flashcards} that are visual representations to {\em capture} the encoded knowledge of a network, as a function of random image patterns. We demonstrate the effectiveness of flashcards in capturing representations and show that they are efficient replay methods for general and task agnostic continual learning setting. Thus, while adapting to a new task, a limited number of constructed flashcards, help to prevent catastrophic forgetting of the previously learned tasks. Most interestingly, such flashcards neither require external memory storage nor need to be accumulated over multiple tasks and only need to be constructed just before learning the subsequent new task, irrespective of the number of tasks trained before and are hence task agnostic. We first demonstrate the efficacy of flashcards in capturing knowledge representation from a trained network, and empirically validate the efficacy of flashcards on a variety of continual learning tasks: continual unsupervised reconstruction, continual denoising, and new-instance learning classification, using a number of heterogeneous benchmark datasets. These studies also indicate that continual learning algorithms with flashcards as the replay strategy perform better than other state-of-the-art replay methods, and exhibits on par performance with the best possible baseline using coreset sampling, with the least additional computational complexity and storage.

112.Generative Adversarial Networks for Automatic Polyp Segmentation ⬇️

This paper aims to contribute in bench-marking the automatic polyp segmentation problem using generative adversarial networks framework. Perceiving the problem as an image-to-image translation task, conditional generative adversarial networks are utilized to generate masks conditioned by the images as inputs. Both generator and discriminator are convolution neural networks based. The model achieved 0.4382 on Jaccard index and 0.611 as F2 score.

113.HI-Net: Hyperdense Inception 3D UNet for Brain Tumor Segmentation ⬇️

The brain tumor segmentation task aims to classify tissue into the whole tumor (WT), tumor core (TC), and enhancing tumor (ET) classes using multimodel MRI images. Quantitative analysis of brain tumors is critical for clinical decision making. While manual segmentation is tedious, time-consuming, and subjective, this task is at the same time very challenging to automatic segmentation methods. Thanks to the powerful learning ability, convolutional neural networks (CNNs), mainly fully convolutional networks, have shown promising brain tumor segmentation. This paper further boosts the performance of brain tumor segmentation by proposing hyperdense inception 3D UNet (HI-Net), which captures multi-scale information by stacking factorization of 3D weighted convolutional layers in the residual inception block. We use hyper dense connections among factorized convolutional layers to extract more contexual information, with the help of features reusability. We use a dice loss function to cope with class imbalances. We validate the proposed architecture on the multi-modal brain tumor segmentation challenges (BRATS) 2020 testing dataset. Preliminary results on the BRATS 2020 testing set show that achieved by our proposed approach, the dice (DSC) scores of ET, WT, and TC are 0.79457, 0.87494, and 0.83712, respectively.

114.Sampling Training Data for Continual Learning Between Robots and the Cloud ⬇️

Today's robotic fleets are increasingly measuring high-volume video and LIDAR sensory streams, which can be mined for valuable training data, such as rare scenes of road construction sites, to steadily improve robotic perception models. However, re-training perception models on growing volumes of rich sensory data in central compute servers (or the "cloud") places an enormous time and cost burden on network transfer, cloud storage, human annotation, and cloud computing resources. Hence, we introduce HarvestNet, an intelligent sampling algorithm that resides on-board a robot and reduces system bottlenecks by only storing rare, useful events to steadily improve perception models re-trained in the cloud. HarvestNet significantly improves the accuracy of machine-learning models on our novel dataset of road construction sites, field testing of self-driving cars, and streaming face recognition, while reducing cloud storage, dataset annotation time, and cloud compute time by between 65.7-81.3%. Further, it is between 1.05-2.58x more accurate than baseline algorithms and scalably runs on embedded deep learning hardware. We provide a suite of compute-efficient perception models for the Google Edge Tensor Processing Unit (TPU), an extended technical report, and a novel video dataset to the research community at this https URL.

115.Learning Consistent Deep Generative Models from Sparse Data via Prediction Constraints ⬇️

We develop a new framework for learning variational autoencoders and other deep generative models that balances generative and discriminative goals. Our framework optimizes model parameters to maximize a variational lower bound on the likelihood of observed data, subject to a task-specific prediction constraint that prevents model misspecification from leading to inaccurate predictions. We further enforce a consistency constraint, derived naturally from the generative model, that requires predictions on reconstructed data to match those on the original data. We show that these two contributions -- prediction constraints and consistency constraints -- lead to promising image classification performance, especially in the semi-supervised scenario where category labels are sparse but unlabeled data is plentiful. Our approach enables advances in generative modeling to directly boost semi-supervised classification performance, an ability we demonstrate by augmenting deep generative models with latent variables capturing spatial transformations.