20200408.md

ArXiv cs.CV --Wed, 8 Apr 2020

1.Feature Pyramid Grids ⬇️

Feature pyramid networks have been widely adopted in the object detection literature to improve feature representations for better handling of variations in scale. In this paper, we present Feature Pyramid Grids (FPG), a deep multi-pathway feature pyramid, that represents the feature scale-space as a regular grid of parallel bottom-up pathways which are fused by multi-directional lateral connections. FPG can improve single-pathway feature pyramid networks by significantly increasing its performance at similar computation cost, highlighting importance of deep pyramid representations. In addition to its general and uniform structure, over complicated structures that have been found with neural architecture search, it also compares favorably against such approaches without relying on search. We hope that FPG with its uniform and effective nature can serve as a strong component for future work in object recognition.

2.Event Based, Near Eye Gaze Tracking Beyond 10,000Hz ⬇️

Fast and accurate eye tracking is crucial for many applications. Current camera-based eye tracking systems, however, are fundamentally limited by their bandwidth, forcing a tradeoff between image resolution and framerate, i.e. between latency and update rate. Here, we propose a hybrid frame-event-based near-eye gaze tracking system offering update rates beyond 10,000 Hz with an accuracy that matches that of high-end desktop-mounted commercial eye trackers when evaluated in the same conditions. Our system builds on emerging event cameras that simultaneously acquire regularly sampled frames and adaptively sampled events. We develop an online 2D pupil fitting method that updates a parametric model every one or few events. Moreover, we propose a polynomial regressor for estimating the gaze vector from the parametric pupil model in real time. Using the first hybrid frame-event gaze dataset, which will be made public, we demonstrate that our system achieves accuracies of 0.45 degrees -- 1.75 degrees for fields of view ranging from 45 degrees to 98 degrees.

3.Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation ⬇️

In this paper, we propose a novel system named Disp R-CNN for 3D object detection from stereo images. Many recent works solve this problem by first recovering a point cloud with disparity estimation and then apply a 3D detector. The disparity map is computed for the entire image, which is costly and fails to leverage category-specific prior. In contrast, we design an instance disparity estimation network (iDispNet) that predicts disparity only for pixels on objects of interest and learns a category-specific shape prior for more accurate disparity estimation. To address the challenge from scarcity of disparity annotation in training, we propose to use a statistical shape model to generate dense disparity pseudo-ground-truth without the need of LiDAR point clouds, which makes our system more widely applicable. Experiments on the KITTI dataset show that, even when LiDAR ground-truth is not available at training time, Disp R-CNN achieves competitive performance and outperforms previous state-of-the-art methods by 20% in terms of average precision.

4.Temporal Pyramid Network for Action Recognition ⬇️

Visual tempo characterizes the dynamics and the temporal scale of an action. Modeling such visual tempos of different actions facilitates their recognition. Previous works often capture the visual tempo through sampling raw videos at multiple rates and constructing an input-level frame pyramid, which usually requires a costly multi-branch network to handle. In this work we propose a generic Temporal Pyramid Network (TPN) at the feature-level, which can be flexibly integrated into 2D or 3D backbone networks in a plug-and-play manner. Two essential components of TPN, the source of features and the fusion of features, form a feature hierarchy for the backbone so that it can capture action instances at various tempos. TPN also shows consistent improvements over other challenging baselines on several action recognition datasets. Specifically, when equipped with TPN, the 3D ResNet-50 with dense sampling obtains a 2% gain on the validation set of Kinetics-400. A further analysis also reveals that TPN gains most of its improvements on action classes that have large variances in their visual tempos, validating the effectiveness of TPN.

5.Unsupervised Person Re-identification via Softened Similarity Learning ⬇️

Person re-identification (re-ID) is an important topic in computer vision. This paper studies the unsupervised setting of re-ID, which does not require any labeled information and thus is freely deployed to new scenarios. There are very few studies under this setting, and one of the best approach till now used iterative clustering and classification, so that unlabeled images are clustered into pseudo classes for a classifier to get trained, and the updated features are used for clustering and so on. This approach suffers two problems, namely, the difficulty of determining the number of clusters, and the hard quantization loss in clustering. In this paper, we follow the iterative training mechanism but discard clustering, since it incurs loss from hard quantization, yet its only product, image-level similarity, can be easily replaced by pairwise computation and a softened classification task. With these improvements, our approach becomes more elegant and is more robust to hyper-parameter changes. Experiments on two image-based and video-based datasets demonstrate state-of-the-art performance under the unsupervised re-ID setting.

6.Dense Regression Network for Video Grounding ⬇️

We address the problem of video grounding from natural language queries. The key challenge in this task is that one training video might only contain a few annotated starting/ending frames that can be used as positive examples for model training. Most conventional approaches directly train a binary classifier using such imbalance data, thus achieving inferior results. The key idea of this paper is to use the distances between the frame within the ground truth and the starting (ending) frame as dense supervisions to improve the video grounding accuracy. Specifically, we design a novel dense regression network (DRN) to regress the distances from each frame to the starting (ending) frame of the video segment described by the query. We also propose a simple but effective IoU regression head module to explicitly consider the localization quality of the grounding results (i.e., the IoU between the predicted location and the ground truth). Experimental results show that our approach significantly outperforms state-of-the-arts on three datasets (i.e., Charades-STA, ActivityNet-Captions, and TACoS).

7.Deep Multi-Shot Network for modelling Appearance Similarity in Multi-Person Tracking applications ⬇️

The automatization of Multi-Object Tracking becomes a demanding task in real unconstrained scenarios, where the algorithms have to deal with crowds, crossing people, occlusions, disappearances and the presence of visually similar individuals. In those circumstances, the data association between the incoming detections and their corresponding identities could miss some tracks or produce identity switches. In order to reduce these tracking errors, and even their propagation in further frames, this article presents a Deep Multi-Shot neural model for measuring the Degree of Appearance Similarity (MS-DoAS) between person observations. This model provides temporal consistency to the individuals' appearance representation, and provides an affinity metric to perform frame-by-frame data association, allowing online tracking. The model has been deliberately trained to be able to manage the presence of previous identity switches and missed observations in the handled tracks. With that purpose, a novel data generation tool has been designed to create training tracklets that simulate such situations. The model has demonstrated a high capacity to discern when a new observation corresponds to a certain track, achieving a classification accuracy of 97% in a hard test that simulates tracks with previous mistakes. Moreover, the tracking efficiency of the model in a Surveillance application has been demonstrated by integrating that into the frame-by-frame association of a Tracking-by-Detection algorithm.

8.Bayesian aggregation improves traditional single image crop classification approaches ⬇️

Machine learning (ML) methods and neural networks (NN) are widely implemented for crop types recognition and classification based on satellite images. However, most of these studies use several multi-temporal images which could be inapplicable for cloudy regions. We present a comparison between the classical ML approaches and U-Net NN for classifying crops with a single satellite image. The results show the advantages of using field-wise classification over pixel-wise approach. We first used a Bayesian aggregation for field-wise classification and improved on 1.5% results between majority voting aggregation. The best result for single satellite image crop classification is achieved for gradient boosting with an overall accuracy of 77.4% and macro F1-score 0.66.

9.Hierarchical Image Classification using Entailment Cone Embeddings ⬇️

Image classification has been studied extensively, but there has been limited work in using unconventional, external guidance other than traditional image-label pairs for training. We present a set of methods for leveraging information about the semantic hierarchy embedded in class labels. We first inject label-hierarchy knowledge into an arbitrary CNN-based classifier and empirically show that availability of such external semantic information in conjunction with the visual semantics from images boosts overall performance. Taking a step further in this direction, we model more explicitly the label-label and label-image interactions using order-preserving embeddings governed by both Euclidean and hyperbolic geometries, prevalent in natural language, and tailor them to hierarchical image classification and representation learning. We empirically validate all the models on the hierarchical ETHEC dataset.

10.Learning Formation of Physically-Based Face Attributes ⬇️

Based on a combined data set of 4000 high resolution facial scans, we introduce a non-linear morphable face model, capable of producing multifarious face geometry of pore-level resolution, coupled with material attributes for use in physically-based rendering. We aim to maximize the variety of identities, while increasing the robustness of correspondence between unique components, including middle-frequency geometry, albedo maps, specular intensity maps and high-frequency displacement details. Our deep learning based generative model learns to correlate albedo and geometry, which ensures the anatomical correctness of the generated assets. We demonstrate potential use of our generative model for novel identity generation, model fitting, interpolation, animation, high fidelity data visualization, and low-to-high resolution data domain transferring. We hope the release of this generative model will encourage further cooperation between all graphics, vision, and data focused professionals, while demonstrating the cumulative value of every individual's complete biometric profile.

11.Simultaneous Learning from Human Pose and Object Cues for Real-Time Activity Recognition ⬇️

Real-time human activity recognition plays an essential role in real-world human-centered robotics applications, such as assisted living and human-robot collaboration. Although previous methods based on skeletal data to encode human poses showed promising results on real-time activity recognition, they lacked the capability to consider the context provided by objects within the scene and in use by the humans, which can provide a further discriminant between human activity categories. In this paper, we propose a novel approach to real-time human activity recognition, through simultaneously learning from observations of both human poses and objects involved in the human activity. We formulate human activity recognition as a joint optimization problem under a unified mathematical framework, which uses a regression-like loss function to integrate human pose and object cues and defines structured sparsity-inducing norms to identify discriminative body joints and object attributes. To evaluate our method, we perform extensive experiments on two benchmark datasets and a physical robot in a home assistance setting. Experimental results have shown that our method outperforms previous methods and obtains real-time performance for human activity recognition with a processing speed of 10^4 Hz.

12.Strategies for Robust Image Classification ⬇️

In this work we evaluate the impact of digitally altered images on the performance of artificial neural networks. We explore factors that negatively affect the ability of an image classification model to produce consistent and accurate results. A model's ability to classify is negatively influenced by alterations to images as a result of digital abnormalities or changes in the physical environment. The focus of this paper is to discover and replicate scenarios that modify the appearance of an image and evaluate them on state-of-the-art machine learning models. Our contributions present various training techniques that enhance a model's ability to generalize and improve robustness against these alterations.

13.RSS-Net: Weakly-Supervised Multi-Class Semantic Segmentation with FMCW Radar ⬇️

This paper presents an efficient annotation procedure and an application thereof to end-to-end, rich semantic segmentation of the sensed environment using FMCW scanning radar. We advocate radar over the traditional sensors used for this task as it operates at longer ranges and is substantially more robust to adverse weather and illumination conditions. We avoid laborious manual labelling by exploiting the largest radar-focused urban autonomy dataset collected to date, correlating radar scans with RGB cameras and LiDAR sensors, for which semantic segmentation is an already consolidated procedure. The training procedure leverages a state-of-the-art natural image segmentation system which is publicly available and as such, in contrast to previous approaches, allows for the production of copious labels for the radar stream by incorporating four camera and two LiDAR streams. Additionally, the losses are computed taking into account labels to the radar sensor horizon by accumulating LiDAR returns along a pose-chain ahead and behind of the current vehicle position. Finally, we present the network with multi-channel radar scan inputs in order to deal with ephemeral and dynamic scene objects.

14.Nonparametric Data Analysis on the Space of Perceived Colors ⬇️

Moving around in a 3D world, requires the visual system of a living individual to rely on three channels of image recognition, which is done through three types of retinal cones. Newton, Grasmann, Helmholz and Schr$\ddot{o}$dinger laid down the basic assumptions needed to understand colored vision. Such concepts were furthered by Resnikoff, who imagined the space of perceived colors as a 3D homogeneous space.
This article is concerned with perceived colors regarded as random objects on a Resnikoff 3D homogeneous space model. Two applications to color differentiation in machine vision are illustrated for the proposed statistical methodology, applied to the Euclidean model for perceived colors.

15.MNEW: Multi-domain Neighborhood Embedding and Weighting for Sparse Point Clouds Segmentation ⬇️

Point clouds have been widely adopted in 3D semantic scene understanding. However, point clouds for typical tasks such as 3D shape segmentation or indoor scenario parsing are much denser than outdoor LiDAR sweeps for the application of autonomous driving perception. Due to the spatial property disparity, many successful methods designed for dense point clouds behave depreciated effectiveness on the sparse data. In this paper, we focus on the semantic segmentation task of sparse outdoor point clouds. We propose a new method called MNEW, including multi-domain neighborhood embedding, and attention weighting based on their geometry distance, feature similarity, and neighborhood sparsity. The network architecture inherits PointNet which directly process point clouds to capture pointwise details and global semantics, and is improved by involving multi-scale local neighborhoods in static geometry domain and dynamic feature space. The distance/similarity attention and sparsity-adapted weighting mechanism of MNEW enable its capability for a wide range of data sparsity distribution. With experiments conducted on virtual and real KITTI semantic datasets, MNEW achieves the top performance for sparse point clouds, which is important to the application of LiDAR-based automated driving perception.

16.Differential 3D Facial Recognition: Adding 3D to Your State-of-the-Art 2D Method ⬇️

Active illumination is a prominent complement to enhance 2D face recognition and make it more robust, e.g., to spoofing attacks and low-light conditions. In the present work we show that it is possible to adopt active illumination to enhance state-of-the-art 2D face recognition approaches with 3D features, while bypassing the complicated task of 3D reconstruction. The key idea is to project over the test face a high spatial frequency pattern, which allows us to simultaneously recover real 3D information plus a standard 2D facial image. Therefore, state-of-the-art 2D face recognition solution can be transparently applied, while from the high frequency component of the input image, complementary 3D facial features are extracted. Experimental results on ND-2006 dataset show that the proposed ideas can significantly boost face recognition performance and dramatically improve the robustness to spoofing attacks.

17.Attribution in Scale and Space ⬇️

We study the attribution problem [28] for deep networks applied to perception tasks. For vision tasks, attribution techniques attribute the prediction of a network to the pixels of the input image. We propose a new technique called \emph{Blur Integrated Gradients}. This technique has several advantages over other methods. First, it can tell at what scale a network recognizes an object. It produces scores in the scale/frequency dimension, that we find captures interesting phenomena. Second, it satisfies the scale-space axioms [14], which imply that it employs perturbations that are free of artifact. We therefore produce explanations that are cleaner and consistent with the operation of deep networks. Third, it eliminates the need for a 'baseline' parameter for Integrated Gradients [31] for perception tasks. This is desirable because the choice of baseline has a significant effect on the explanations. We compare the proposed technique against previous techniques and demonstrate application on three tasks: ImageNet object recognition, Diabetic Retinopathy prediction, and AudioSet audio event identification.

18.Error-Corrected Margin-Based Deep Cross-Modal Hashing for Facial Image Retrieval ⬇️

Cross-modal hashing facilitates mapping of heterogeneous multimedia data into a common Hamming space, which can beutilized for fast and flexible retrieval across different modalities. In this paper, we propose a novel cross-modal hashingarchitecture-deep neural decoder cross-modal hashing (DNDCMH), which uses a binary vector specifying the presence of certainfacial attributes as an input query to retrieve relevant face images from a database. The DNDCMH network consists of two separatecomponents: an attribute-based deep cross-modal hashing (ADCMH) module, which uses a margin (m)-based loss function toefficiently learn compact binary codes to preserve similarity between modalities in the Hamming space, and a neural error correctingdecoder (NECD), which is an error correcting decoder implemented with a neural network. The goal of NECD network in DNDCMH isto error correct the hash codes generated by ADCMH to improve the retrieval efficiency. The NECD network is trained such that it hasan error correcting capability greater than or equal to the margin (m) of the margin-based loss function. This results in NECD cancorrect the corrupted hash codes generated by ADCMH up to the Hamming distance of m. We have evaluated and comparedDNDCMH with state-of-the-art cross-modal hashing methods on standard datasets to demonstrate the superiority of our method.

19.Composition of Saliency Metrics for Channel Pruning with a Myopic Oracle ⬇️

The computation and memory needed for Convolutional Neural Network (CNN) inference can be reduced by pruning weights from the trained network. Pruning is guided by a pruning saliency, which heuristically approximates the change in the loss function associated with the removal of specific weights. Many pruning signals have been proposed, but the performance of each heuristic depends on the particular trained network. This leaves the data scientist with a difficult choice. When using any one saliency metric for the entire pruning process, we run the risk of the metric assumptions being invalidated, leading to poor decisions being made by the metric. Ideally we could combine the best aspects of different saliency metrics. However, despite an extensive literature review, we are unable to find any prior work on composing different saliency metrics. The chief difficulty lies in combining the numerical output of different saliency metrics, which are not directly comparable.
We propose a method to compose several primitive pruning saliencies, to exploit the cases where each saliency measure does well. Our experiments show that the composition of saliencies avoids many poor pruning choices identified by individual saliencies. In most cases our method finds better selections than even the best individual pruning saliency.

20.Robust Self-Supervised Convolutional Neural Network for Subspace Clustering and Classification ⬇️

Insufficient capability of existing subspace clustering methods to handle data coming from nonlinear manifolds, data corruptions, and out-of-sample data hinders their applicability to address real-world clustering and classification problems. This paper proposes the robust formulation of the self-supervised convolutional subspace clustering network ($S^2$ConvSCN) that incorporates the fully connected (FC) layer and, thus, it is capable for handling out-of-sample data by classifying them using a softmax classifier. $S^2$ConvSCN clusters data coming from nonlinear manifolds by learning the linear self-representation model in the feature space. Robustness to data corruptions is achieved by using the correntropy induced metric (CIM) of the error. Furthermore, the block-diagonal (BD) structure of the representation matrix is enforced explicitly through BD regularization. In a truly unsupervised training environment, Robust $S^2$ConvSCN outperforms its baseline version by a significant amount for both seen and unseen data on four well-known datasets. Arguably, such an ablation study has not been reported before.

21.Improving BPSO-based feature selection applied to offline WI handwritten signature verification through overfitting control ⬇️

This paper investigates the presence of overfitting when using Binary Particle Swarm Optimization (BPSO) to perform the feature selection in a context of Handwritten Signature Verification (HSV). SigNet is a state of the art Deep CNN model for feature representation in the HSV context and contains 2048 dimensions. Some of these dimensions may include redundant information in the dissimilarity representation space generated by the dichotomy transformation (DT) used by the writer-independent (WI) approach. The analysis is carried out on the GPDS-960 dataset. Experiments demonstrate that the proposed method is able to control overfitting during the search for the most discriminant representation.

22.A white-box analysis on the writer-independent dichotomy transformation applied to offline handwritten signature verification ⬇️

High number of writers, small number of training samples per writer with high intra-class variability and heavily imbalanced class distributions are among the challenges and difficulties of the offline Handwritten Signature Verification (HSV) problem. A good alternative to tackle these issues is to use a writer-independent (WI) framework. In WI systems, a single model is trained to perform signature verification for all writers from a dissimilarity space generated by the dichotomy transformation. Among the advantages of this framework is its scalability to deal with some of these challenges and its ease in managing new writers, and hence of being used in a transfer learning context. In this work, we present a white-box analysis of this approach highlighting how it handles the challenges, the dynamic selection of references through fusion function, and its application for transfer learning. All the analyses are carried out at the instance level using the instance hardness (IH) measure. The experimental results show that, using the IH analysis, we were able to characterize "good" and "bad" quality skilled forgeries as well as the frontier region between positive and negative samples. This enables futures investigations on methods for improving discrimination between genuine signatures and skilled forgeries by considering these characterizations.

23.Knife and Threat Detectors ⬇️

Despite rapid advances in image-based machine learning, the threat identification of a knife wielding attacker has not garnered substantial academic attention. This relative research gap appears less understandable given the high knife assault rate (>100,000 annually) and the increasing availability of public video surveillance to analyze and forensically document. We present three complementary methods for scoring automated threat identification using multiple knife image datasets, each with the goal of narrowing down possible assault intentions while minimizing misidentifying false positives and risky false negatives. To alert an observer to the knife-wielding threat, we test and deploy classification built around MobileNet in a sparse and pruned neural network with a small memory requirement (< 2.2 megabytes) and 95% test accuracy. We secondly train a detection algorithm (MaskRCNN) to segment the hand from the knife in a single image and assign probable certainty to their relative location. This segmentation accomplishes both localization with bounding boxes but also relative positions to infer overhand threats. A final model built on the PoseNet architecture assigns anatomical waypoints or skeletal features to narrow the threat characteristics and reduce misunderstood intentions. We further identify and supplement existing data gaps that might blind a deployed knife threat detector such as collecting innocuous hand and fist images as important negative training sets. When automated on commodity hardware and software solutions one original research contribution is this systematic survey of timely and readily available image-based alerts to task and prioritize crime prevention countermeasures prior to a tragic outcome.

24.A Machine Learning Based Framework for the Smart Healthcare Monitoring ⬇️

In this paper, we propose a novel framework for the smart healthcare system, where we employ the compressed sensing (CS) and the combination of the state-of-the-art machine learning based denoiser as well as the alternating direction of method of multipliers (ADMM) structure. This integration significantly simplifies the software implementation for the lowcomplexity encoder, thanks to the modular structure of ADMM. Furthermore, we focus on detecting fall down actions from image streams. Thus, teh primary purpose of thus study is to reconstruct the image as visibly clear as possible and hence it helps the detection step at the trained classifier. For this efficient smart health monitoring framework, we employ the trained binary convolutional neural network (CNN) classifier for the fall-action classifier, because this scheme is a part of surveillance scenario. In this scenario, we deal with the fallimages, thus, we compress, transmit and reconstruct the fallimages. Experimental results demonstrate the impacts of network parameters and the significant performance gain of the proposal compared to traditional methods.

25.Deep learning approaches in food recognition ⬇️

Automatic image-based food recognition is a particularly challenging task. Traditional image analysis approaches have achieved low classification accuracy in the past, whereas deep learning approaches enabled the identification of food types and their ingredients. The contents of food dishes are typically deformable objects, usually including complex semantics, which makes the task of defining their structure very difficult. Deep learning methods have already shown very promising results in such challenges, so this chapter focuses on the presentation of some popular approaches and techniques applied in image-based food recognition. The three main lines of solutions, namely the design from scratch, the transfer learning and the platform-based approaches, are outlined, particularly for the task at hand, and are tested and compared to reveal the inherent strengths and weaknesses. The chapter is complemented with basic background material, a section devoted to the relevant datasets that are crucial in light of the empirical approaches adopted, and some concluding remarks that underline the future directions.

26.Inclusive GAN: Improving Data and Minority Coverage in Generative Models ⬇️

Generative Adversarial Networks (GANs) have brought about rapid progress towards generating photorealistic images. Yet the equitable allocation of their modeling capacity among subgroups has received less attention, which could lead to potential biases against underrepresented minorities if left uncontrolled. In this work, we first formalize the problem of minority inclusion as one of data coverage, and then propose to improve data coverage by harmonizing adversarial training with reconstructive generation. The experiments show that our method outperforms the existing state-of-the-art methods in terms of data coverage on both seen and unseen data. We develop an extension that allows explicit control over the minority subgroups that the model should ensure to include, and validate its effectiveness at little compromise from the overall performance on the entire dataset. Code, models, and supplemental videos are available at GitHub.

27.An Image Labeling Tool and Agricultural Dataset for Deep Learning ⬇️

We introduce a labeling tool and dataset aimed to facilitate computer vision research in agriculture. The annotation tool introduces novel methods for labeling with a variety of manual, semi-automatic, and fully-automatic tools. The dataset includes original images collected from commercial greenhouses, images from PlantVillage, and images from Google Images. Images were annotated with segmentations for foreground leaf, fruit, and stem instances, and diseased leaf area. Labels were in an extended COCO format. In total the dataset contained 10k tomatoes, 7k leaves, 2k stems, and 2k diseased leaf annotations.

28.Automatic Generation of Chinese Handwriting via Fonts Style Representation Learning ⬇️

In this paper, we propose and end-to-end deep Chinese font generation system. This system can generate new style fonts by interpolation of latent style-related embeding variables that could achieve smooth transition between different style. Our method is simpler and more effective than other methods, which will help to improve the font design efficiency

29.Multiform Fonts-to-Fonts Translation via Style and Content Disentangled Representations of Chinese Character ⬇️

This paper mainly discusses the generation of personalized fonts as the problem of image style transfer. The main purpose of this paper is to design a network framework that can extract and recombine the content and style of the characters. These attempts can be used to synthesize the entire set of fonts with only a small amount of characters. The paper combines various depth networks such as Convolutional Neural Network, Multi-layer Perceptron and Residual Network to find the optimal model to extract the features of the fonts character. The result shows that those characters we have generated is very close to real characters, using Structural Similarity index and Peak Signal-to-Noise Ratio evaluation criterions.

30.An End-to-End Approach for Recognition of Modern and Historical Handwritten Numeral Strings ⬇️

An end-to-end solution for handwritten numeral string recognition is proposed, in which the numeral string is considered as composed of objects automatically detected and recognized by a YoLo-based model. The main contribution of this paper is to avoid heuristic-based methods for string preprocessing and segmentation, the need for task-oriented classifiers, and also the use of specific constraints related to the string length. A robust experimental protocol based on several numeral string datasets, including one composed of historical documents, has shown that the proposed method is a feasible end-to-end solution for numeral string recognition. Besides, it reduces the complexity of the string recognition task considerably since it drops out classical steps, in special preprocessing, segmentation, and a set of classifiers devoted to strings with a specific length.

31.Predict the model of a camera ⬇️

In this work we address the problem of predicting the model of a camera based on the content of their photographs. We use two set of features, one set consist in properties extracted from a Discrete Wavelet Domain (DWD) obtained by applying a 4 level Fast Wavelet Decomposition of the images, and a second set are Local Binary Patterns (LBP) features from the after filter noise of images. The algorithms used for classification were Logistic regression, K-NN and Artificial Neural Networks

32.FusedProp: Towards Efficient Training of Generative Adversarial Networks ⬇️

Generative adversarial networks (GANs) are capable of generating strikingly realistic samples but state-of-the-art GANs can be extremely computationally expensive to train. In this paper, we propose the fused propagation (FusedProp) algorithm which can be used to efficiently train the discriminator and the generator of common GANs simultaneously using only one forward and one backward propagation. We show that FusedProp achieves 1.49 times the training speed compared to the conventional training of GANs, although further studies are required to improve its stability. By reporting our preliminary results and open-sourcing our implementation, we hope to accelerate future research on the training of GANs.

33.Streaming Networks: Increase Noise Robustness and Filter Diversity via Hard-wired and Input-induced Sparsity ⬇️

The CNNs have achieved a state-of-the-art performance in many applications. Recent studies illustrate that CNN's recognition accuracy drops drastically if images are noise corrupted. We focus on the problem of robust recognition accuracy of noise-corrupted images. We introduce a novel network architecture called Streaming Networks. Each stream is taking a certain intensity slice of the original image as an input, and stream parameters are trained independently. We use network capacity, hard-wired and input-induced sparsity as the dimensions for experiments. The results indicate that only the presence of both hard-wired and input-induces sparsity enables robust noisy image recognition. Streaming Nets is the only architecture which has both types of sparsity and exhibits higher robustness to noise. Finally, to illustrate increase in filter diversity we illustrate that a distribution of filter weights of the first conv layer gradually approaches uniform distribution as the degree of hard-wired and domain-induced sparsity and capacities increases.

34.Cascaded Refinement Network for Point Cloud Completion ⬇️

Point clouds are often sparse and incomplete. Existing shape completion methods are incapable of generating details of objects or learning the complex point distributions. To this end, we propose a cascaded refinement network together with a coarse-to-fine strategy to synthesize the detailed object shapes. Considering the local details of partial input with the global shape information together, we can preserve the existing details in the incomplete point set and generate the missing parts with high fidelity. We also design a patch discriminator that guarantees every local area has the same pattern with the ground truth to learn the complicated point distribution. Quantitative and qualitative experiments on different datasets show that our method achieves superior results compared to existing state-of-the-art approaches on the 3D point cloud completion task. Our source code is available at this https URL.

35.Towards Efficient Unconstrained Palmprint Recognition via Deep Distillation Hashing ⬇️

Deep palmprint recognition has become an emerging issue with great potential for personal authentication on handheld and wearable consumer devices. Previous studies of palmprint recognition are mainly based on constrained datasets collected by dedicated devices in controlled environments, which has to reduce the flexibility and convenience. In addition, general deep palmprint recognition algorithms are often too heavy to meet the real-time requirements of embedded system. In this paper, a new palmprint benchmark is established, which consists of more than 20,000 images collected by 5 brands of smart phones in an unconstrained manner. Each image has been manually labeled with 14 key points for region of interest (ROI) extraction. Further, the approach called Deep Distillation Hashing (DDH) is proposed as benchmark for efficient deep palmprint recognition. Palmprint images are converted to binary codes to improve the efficiency of feature matching. Derived from knowledge distillation, novel distillation loss functions are constructed to compress deep model to further improve the efficiency of feature extraction on light network. Comprehensive experiments are conducted on both constrained and unconstrained palmprint databases. Using DDH, the accuracy of palmprint identification can be increased by up to 11.37%, and the Equal Error Rate (EER) of palmprint verification can be reduced by up to 3.11%. The results indicate the feasibility of our database, and DDH can outperform other baselines to achieve the state-of-the-art performance. The collected dataset and related source codes are publicly available at this http URL.

36.Pyramid Focusing Network for mutation prediction and classification in CT images ⬇️

Predicting the mutation status of genes in tumors is of great clinical significance. Recent studies have suggested that certain mutations may be noninvasively predicted by studying image features of the tumors from Computed Tomography (CT) data. Currently, this kind of image feature identification method mainly relies on manual processing to extract generalized image features alone or machine processing without considering the morphological differences of the tumor itself, which makes it difficult to achieve further breakthroughs. In this paper, we propose a pyramid focusing network (PFNet) for mutation prediction and classification based on CT images. Firstly, we use Space Pyramid Pooling to collect semantic cues in feature maps from multiple scales according to the observation that the shape and size of the tumors are varied.Secondly, we improve the loss function based on the consideration that the features required for proper mutation detection are often not obvious in cross-sections of tumor edges, which raises more attention to these hard examples in the network. Finally, we devise a training scheme based on data augmentation to enhance the generalization ability of networks. Extensively verified on clinical gastric CT datasets of 20 testing volumes with 63648 CT images, our method achieves the accuracy of 94.90% in predicting the HER-2 genes mutation status of at the CT image.

37.Super-resolution of clinical CT volumes with modified CycleGAN using micro CT volumes ⬇️

This paper presents a super-resolution (SR) method with unpaired training dataset of clinical CT and micro CT volumes. For obtaining very detailed information such as cancer invasion from pre-operative clinical CT volumes of lung cancer patients, SR of clinical CT volumes to $\m$}CT level is desired. While most SR methods require paired low- and high- resolution images for training, it is infeasible to obtain paired clinical CT and {\mu}CT volumes. We propose a SR approach based on CycleGAN, which could perform SR on clinical CT into $\mu$CT level. We proposed new loss functions to keep cycle consistency, while training without paired volumes. Experimental results demonstrated that our proposed method successfully performed SR of clinical CT volume of lung cancer patients into $\mu$CT level.

38.SC4D: A Sparse 4D Convolutional Network for Skeleton-Based Action Recognition ⬇️

In this paper, a new perspective is presented for skeleton-based action recognition. Specifically, we regard the skeletal sequence as a spatial-temporal point cloud and voxelize it into a 4-dimensional grid. A novel sparse 4D convolutional network (SC4D) is proposed to directly process the generated 4D grid for high-level perceptions. Without manually designing the hand-crafted transformation rules, it makes better use of the advantages of the convolutional network, resulting in a more concise, general and robust framework for skeletal data. Besides, by processing the space and time simultaneously, it largely keeps the spatial-temporal consistency of the skeletal data, and thus brings better expressiveness. Moreover, with the help of the sparse tensor, it can be efficiently executed with less computations. To verify the superiority of SC4D, extensive experiments are conducted on two challenging datasets, namely, NTU-RGBD and SHREC, where SC4D achieves state-of-the-art performance on both of them.

39.Hierarchical Opacity Propagation for Image Matting ⬇️

Natural image matting is a fundamental problem in computational photography and computer vision. Deep neural networks have seen the surge of successful methods in natural image matting in recent years. In contrast to traditional propagation-based matting methods, some top-tier deep image matting approaches tend to perform propagation in the neural network implicitly. A novel structure for more direct alpha matte propagation between pixels is in demand. To this end, this paper presents a hierarchical opacity propagation (HOP) matting method, where the opacity information is propagated in the neighborhood of each point at different semantic levels. The hierarchical structure is based on one global and multiple local propagation blocks. With the HOP structure, every feature point pair in high-resolution feature maps will be connected based on the appearance of input image. We further propose a scale-insensitive positional encoding tailored for image matting to deal with the unfixed size of input image and introduce the random interpolation augmentation into image matting. Extensive experiments and ablation study show that HOP matting is capable of outperforming state-of-the-art matting methods.

40.Motion-supervised Co-Part Segmentation ⬇️

Recent co-part segmentation methods mostly operate in a supervised learning setting, which requires a large amount of annotated data for training. To overcome this limitation, we propose a self-supervised deep learning method for co-part segmentation. Differently from previous works, our approach develops the idea that motion information inferred from videos can be leveraged to discover meaningful object parts. To this end, our method relies on pairs of frames sampled from the same video. The network learns to predict part segments together with a representation of the motion between two frames, which permits reconstruction of the target image. Through extensive experimental evaluation on publicly available video sequences we demonstrate that our approach can produce improved segmentation maps with respect to previous self-supervised co-part segmentation approaches.

41.Utilising Prior Knowledge for Visual Navigation: Distil and Adapt ⬇️

We, as humans, can impeccably navigate to localise a target object, even in an unseen environment. We argue that this impressive ability is largely due to incorporation of \emph{prior knowledge} (or experience) and \emph{visual cues}--that current visual navigation approaches lack. In this paper, we propose to use externally learned prior knowledge of object relations, which is integrated to our model via constructing a neural graph. To combine appropriate assessment of the states and the prior (knowledge), we propose to decompose the value function in the actor-critic reinforcement learning algorithm and incorporate the prior in the critic in a novel way that reduces the model complexity and improves model generalisation. Our approach outperforms the current state-of-the-art in AI2THOR visual navigation dataset.

42.Neural Image Inpainting Guided with Descriptive Text ⬇️

Neural image inpainting has achieved promising performance in generating semantically plausible content. Most of the recent works mainly focus on inpainting images depending on vision information, while neglecting the semantic information implied in human languages. To acquire more semantically accurate inpainting images, this paper proposes a novel inpainting model named \textit{N}eural \textit{I}mage Inpainting \textit{G}uided with \textit{D}escriptive \textit{T}ext (NIGDT). First, a dual multi-modal attention mechanism is designed to extract the explicit semantic information about corrupted regions. The mechanism is trained to combine the descriptive text and two complementary images through reciprocal attention maps. Second, an image-text matching loss is designed to enforce the model output following the descriptive text. Its goal is to maximize the semantic similarity of the generated image and the text. Finally, experiments are conducted on two open datasets with captions. Experimental results show that the proposed NIGDT model outperforms all compared models on both quantitative and qualitative comparison. The results also demonstrate that the proposed model can generate images consistent with the guidance text, which provides a flexible way for user-guided inpainting. Our systems and code will be released soon.

43.Multi-Task Learning via Co-Attentive Sharing for Pedestrian Attribute Recognition ⬇️

Learning to predict multiple attributes of a pedestrian is a multi-task learning problem. To share feature representation between two individual task networks, conventional methods like Cross-Stitch and Sluice network learn a linear combination of features or feature subspaces. However, linear combination rules out the complex interdependency between channels. Moreover, spatial information exchanging is less-considered. In this paper, we propose a novel Co-Attentive Sharing (CAS) module which extracts discriminative channels and spatial regions for more effective feature sharing in multi-task learning. The module consists of three branches, which leverage different channels for between-task feature fusing, attention generation and task-specific feature enhancing, respectively. Experiments on two pedestrian attribute recognition datasets show that our module outperforms the conventional sharing units and achieves superior results compared to the state-of-the-art approaches using many metrics.

44.Real-time Classification from Short Event-Camera Streams using Input-filtering Neural ODEs ⬇️

Event-based cameras are novel, efficient sensors inspired by the human vision system, generating an asynchronous, pixel-wise stream of data. Learning from such data is generally performed through heavy preprocessing and event integration into images. This requires buffering of possibly long sequences and can limit the response time of the inference system. In this work, we instead propose to directly use events from a DVS camera, a stream of intensity changes and their spatial coordinates. This sequence is used as the input for a novel \emph{asynchronous} RNN-like architecture, the Input-filtering Neural ODEs (INODE). This is inspired by the dynamical systems and filtering literature. INODE is an extension of Neural ODEs (NODE) that allows for input signals to be continuously fed to the network, like in filtering. The approach naturally handles batches of time series with irregular time-stamps by implementing a batch forward Euler solver. INODE is trained like a standard RNN, it learns to discriminate short event sequences and to perform event-by-event online inference. We demonstrate our approach on a series of classification tasks, comparing against a set of LSTM baselines. We show that, independently of the camera resolution, INODE can outperform the baselines by a large margin on the ASL task and it's on par with a much larger LSTM for the NCALTECH task. Finally, we show that INODE is accurate even when provided with very few events.

45.Adaptive Multiscale Illumination-Invariant Feature Representation for Undersampled Face Recognition ⬇️

This paper presents an novel illumination-invariant feature representation approach used to eliminate the varying illumination affection in undersampled face recognition. Firstly, a new illumination level classification technique based on Singular Value Decomposition (SVD) is proposed to judge the illumination level of input image. Secondly, we construct the logarithm edgemaps feature (LEF) based on lambertian model and local near neighbor feature of the face image, applying to local region within multiple scales. Then, the illumination level is referenced to construct the high performance LEF as well realize adaptive fusion for multiple scales LEFs for the face image, performing JLEF-feature. In addition, the constrain operation is used to remove the useless high-frequency interference, disentangling useful facial feature edges and constructing AJLEF-face. Finally, the effects of the our methods and other state-of-the-art algorithms including deep learning methods are tested on Extended Yale B, CMU PIE, AR as well as our Self-build Driver database (SDB). The experimental results demonstrate that the JLEF-feature and AJLEF-face outperform other related approaches for undersampled face recognition under varying illumination.

46.Predicting Camera Viewpoint Improves Cross-dataset Generalization for 3D Human Pose Estimation ⬇️

Monocular estimation of 3d human pose has attracted increased attention with the availability of large ground-truth motion capture datasets. However, the diversity of training data available is limited and it is not clear to what extent methods generalize outside the specific datasets they are trained on. In this work we carry out a systematic study of the diversity and biases present in specific datasets and its effect on cross-dataset generalization across a compendium of 5 pose datasets. We specifically focus on systematic differences in the distribution of camera viewpoints relative to a body-centered coordinate frame. Based on this observation, we propose an auxiliary task of predicting the camera viewpoint in addition to pose. We find that models trained to jointly predict viewpoint and pose systematically show significantly improved cross-dataset generalization.

47.Human Motion Transfer from Poses in the Wild ⬇️

In this paper, we tackle the problem of human motion transfer, where we synthesize novel motion video for a target person that imitates the movement from a reference video. It is a video-to-video translation task in which the estimated poses are used to bridge two domains. Despite substantial progress on the topic, there exist several problems with the previous methods. First, there is a domain gap between training and testing pose sequences--the model is tested on poses it has not seen during training, such as difficult dancing moves. Furthermore, pose detection errors are inevitable, making the job of the generator harder. Finally, generating realistic pixels from sparse poses is challenging in a single step. To address these challenges, we introduce a novel pose-to-video translation framework for generating high-quality videos that are temporally coherent even for in-the-wild pose sequences unseen during training. We propose a pose augmentation method to minimize the training-test gap, a unified paired and unpaired learning strategy to improve the robustness to detection errors, and two-stage network architecture to achieve superior texture quality. To further boost research on the topic, we build two human motion datasets. Finally, we show the superiority of our approach over the state-of-the-art studies through extensive experiments and evaluations on different datasets.

48.Toward Fine-grained Facial Expression Manipulation ⬇️

Facial expression manipulation, as an image-to-image translation problem, aims at editing facial expression with a given condition. Previous methods edit an input image under the guidance of a discrete emotion label or absolute condition (e.g., facial action units) to possess the desired expression. However, these methods either suffer from changing condition-irrelevant regions or are inefficient to preserve image quality. In this study, we take these two objectives into consideration and propose a novel conditional GAN model. First, we replace continuous absolute condition with relative condition, specifically, relative action units. With relative action units, the generator learns to only transform regions of interest which are specified by non-zero-valued relative AUs, avoiding estimating the current AUs of input image. Second, our generator is built on U-Net architecture and strengthened by multi-scale feature fusion (MSF) mechanism for high-quality expression editing purpose. Extensive experiments on both quantitative and qualitative evaluation demonstrate the improvements of our proposed approach compared with the state-of-the-art expression editing methods.

49.Generative Adversarial Zero-shot Learning via Knowledge Graphs ⬇️

Zero-shot learning (ZSL) is to handle the prediction of those unseen classes that have no labeled training data. Recently, generative methods like Generative Adversarial Networks (GANs) are being widely investigated for ZSL due to their high accuracy, generalization capability and so on. However, the side information of classes used now is limited to text descriptions and attribute annotations, which are in short of semantics of the classes. In this paper, we introduce a new generative ZSL method named KG-GAN by incorporating rich semantics in a knowledge graph (KG) into GANs. Specifically, we build upon Graph Neural Networks and encode KG from two views: class view and attribute view considering the different semantics of KG. With well-learned semantic embeddings for each node (representing a visual category), we leverage GANs to synthesize compelling visual features for unseen classes. According to our evaluation with multiple image classification datasets, KG-GAN can achieve better performance than the state-of-the-art baselines.

50.End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection ⬇️

Reliable and accurate 3D object detection is a necessity for safe autonomous driving. Although LiDAR sensors can provide accurate 3D point cloud estimates of the environment, they are also prohibitively expensive for many settings. Recently, the introduction of pseudo-LiDAR (PL) has led to a drastic reduction in the accuracy gap between methods based on LiDAR sensors and those based on cheap stereo cameras. PL combines state-of-the-art deep neural networks for 3D depth estimation with those for 3D object detection by converting 2D depth map outputs to 3D point cloud inputs. However, so far these two networks have to be trained separately. In this paper, we introduce a new framework based on differentiable Change of Representation (CoR) modules that allow the entire PL pipeline to be trained end-to-end. The resulting framework is compatible with most state-of-the-art networks for both tasks and in combination with PointRCNN improves over PL consistently across all benchmarks -- yielding the highest entry on the KITTI image-based 3D object detection leaderboard at the time of submission. Our code will be made available at this https URL.

51.A Method for Curation of Web-Scraped Face Image Datasets ⬇️

Web-scraped, in-the-wild datasets have become the norm in face recognition research. The numbers of subjects and images acquired in web-scraped datasets are usually very large, with number of images on the millions scale. A variety of issues occur when collecting a dataset in-the-wild, including images with the wrong identity label, duplicate images, duplicate subjects and variation in quality. With the number of images being in the millions, a manual cleaning procedure is not feasible. But fully automated methods used to date result in a less-than-ideal level of clean dataset. We propose a semi-automated method, where the goal is to have a clean dataset for testing face recognition methods, with similar quality across men and women, to support comparison of accuracy across gender. Our approach removes near-duplicate images, merges duplicate subjects, corrects mislabeled images, and removes images outside a defined range of pose and quality. We conduct the curation on the Asian Face Dataset (AFD) and VGGFace2 test dataset. The experiments show that a state-of-the-art method achieves a much higher accuracy on the datasets after they are curated. Finally, we release our cleaned versions of both datasets to the research community.

52.MGGR: MultiModal-Guided Gaze Redirection with Coarse-to-Fine Learning ⬇️

Gaze redirection aims at manipulating a given eye gaze to a desirable direction according to a reference angle and it can be applied to many real life scenarios, such as video-conferencing or taking groups. However, the previous works suffer from two limitations: (1) low-quality generation and (2) low redirection precision. To this end, we propose an innovative MultiModal-Guided Gaze Redirection~(MGGR) framework that fully exploits eye-map images and target angles to adjust a given eye appearance through a designed coarse-to-fine learning. Our contribution is combining the flow-learning and adversarial learning for coarse-to-fine generation. More specifically, the role of the proposed coarse branch with flow field is to rapidly learn the spatial transformation for attaining the warped result with the desired gaze. The proposed fine-grained branch consists of a generator network with conditional residual image learning and a multi-task discriminator to reduce the gap between the warped image and the ground-truth image for recovering the finer texture details. Moreover, we propose leveraging the gazemap for desired angles as an extra guide to further improve the precision of gaze redirection. Extensive experiments on a benchmark dataset show that the proposed method outperforms the state-of-the-art methods in terms of image quality and redirection precision. Further evaluations demonstrate the effectiveness of the proposed coarse-to-fine and gazemap modules.

53.Depth Sensing Beyond LiDAR Range ⬇️

Depth sensing is a critical component of autonomous driving technologies, but today's LiDAR- or stereo camera-based solutions have limited range. We seek to increase the maximum range of self-driving vehicles' depth perception modules for the sake of better safety. To that end, we propose a novel three-camera system that utilizes small field of view cameras. Our system, along with our novel algorithm for computing metric depth, does not require full pre-calibration and can output dense depth maps with practically acceptable accuracy for scenes and objects at long distances not well covered by most commercial LiDARs.

54.Manifold-driven Attention Maps for Weakly Supervised Segmentation ⬇️

Segmentation using deep learning has shown promising directions in medical imaging as it aids in the analysis and diagnosis of diseases. Nevertheless, a main drawback of deep models is that they require a large amount of pixel-level labels, which are laborious and expensive to obtain. To mitigate this problem, weakly supervised learning has emerged as an efficient alternative, which employs image-level labels, scribbles, points, or bounding boxes as supervision. Among these, image-level labels are easier to obtain. However, since this type of annotation only contains object category information, the segmentation task under this learning paradigm is a challenging problem. To address this issue, visual salient regions derived from trained classification networks are typically used. Despite their success to identify important regions on classification tasks, these saliency regions only focus on the most discriminant areas of an image, limiting their use in semantic segmentation. In this work, we propose a manifold driven attention-based network to enhance visual salient regions, thereby improving segmentation accuracy in a weakly supervised setting. Our method generates superior attention maps directly during inference without the need of extra computations. We evaluate the benefits of our approach in the task of segmentation using a public benchmark on skin lesion images. Results demonstrate that our method outperforms the state-of-the-art GradCAM by a margin of ~22% in terms of Dice score.

55.When, Where, and What? A New Dataset for Anomaly Detection in Driving Videos ⬇️

Video anomaly detection (VAD) has been extensively studied. However, research on egocentric traffic videos with dynamic scenes lacks large-scale benchmark datasets as well as effective evaluation metrics. This paper proposes traffic anomaly detection with a \textit{when-where-what} pipeline to detect, localize, and recognize anomalous events from egocentric videos. We introduce a new dataset called Detection of Traffic Anomaly (DoTA) containing 4,677 videos with temporal, spatial, and categorical annotations. A new spatial-temporal area under curve (STAUC) evaluation metric is proposed and used with DoTA. State-of-the-art methods are benchmarked for two VAD-related tasks.Experimental results show STAUC is an effective VAD metric. To our knowledge, DoTA is the largest traffic anomaly dataset to-date and is the first supporting traffic anomaly studies across when-where-what perspectives. Our code and dataset can be found in: this https URL

56.Learning Generative Models of Shape Handles ⬇️

We present a generative model to synthesize 3D shapes as sets of handles -- lightweight proxies that approximate the original 3D shape -- for applications in interactive editing, shape parsing, and building compact 3D representations. Our model can generate handle sets with varying cardinality and different types of handles (Figure 1). Key to our approach is a deep architecture that predicts both the parameters and existence of shape handles, and a novel similarity measure that can easily accommodate different types of handles, such as cuboids or sphere-meshes. We leverage the recent advances in semantic 3D annotation as well as automatic shape summarizing techniques to supervise our approach. We show that the resulting shape representations are intuitive and achieve superior quality than previous state-of-the-art. Finally, we demonstrate how our method can be used in applications such as interactive shape editing, completion, and interpolation, leveraging the latent space learned by our model to guide these tasks. Project page: this http URL.

57.Field-Level Crop Type Classification with k Nearest Neighbors: A Baseline for a New Kenya Smallholder Dataset ⬇️

Accurate crop type maps provide critical information for ensuring food security, yet there has been limited research on crop type classification for smallholder agriculture, particularly in sub-Saharan Africa where risk of food insecurity is highest. Publicly-available ground-truth data such as the newly-released training dataset of crop types in Kenya (Radiant MLHub) are catalyzing this research, but it is important to understand the context of when, where, and how these datasets were obtained when evaluating classification performance and using them as a benchmark across methods. In this paper, we provide context for the new western Kenya dataset which was collected during an atypical 2019 main growing season and demonstrate classification accuracy up to 64% for maize and 70% for cassava using k Nearest Neighbors--a fast, interpretable, and scalable method that can serve as a baseline for future work.

58.Adaptive Fractional Dilated Convolution Network for Image Aesthetics Assessment ⬇️

To leverage deep learning for image aesthetics assessment, one critical but unsolved issue is how to seamlessly incorporate the information of image aspect ratios to learn more robust models. In this paper, an adaptive fractional dilated convolution (AFDC), which is aspect-ratio-embedded, composition-preserving and parameter-free, is developed to tackle this issue natively in convolutional kernel level. Specifically, the fractional dilated kernel is adaptively constructed according to the image aspect ratios, where the interpolation of nearest two integers dilated kernels is used to cope with the misalignment of fractional sampling. Moreover, we provide a concise formulation for mini-batch training and utilize a grouping strategy to reduce computational overhead. As a result, it can be easily implemented by common deep learning libraries and plugged into popular CNN architectures in a computation-efficient manner. Our experimental results demonstrate that our proposed method achieves state-of-the-art performance on image aesthetics assessment over the AVA dataset.

59.LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood ⬇️

Modern face alignment methods have become quite accurate at predicting the locations of facial landmarks, but they do not typically estimate the uncertainty of their predicted locations nor predict whether landmarks are visible. In this paper, we present a novel framework for jointly predicting landmark locations, associated uncertainties of these predicted locations, and landmark visibilities. We model these as mixed random variables and estimate them using a deep network trained with our proposed Location, Uncertainty, and Visibility Likelihood (LUVLi) loss. In addition, we release an entirely new labeling of a large face alignment dataset with over 19,000 face images in a full range of head poses. Each face is manually labeled with the ground-truth locations of 68 landmarks, with the additional information of whether each landmark is unoccluded, self-occluded (due to extreme head poses), or externally occluded. Not only does our joint estimation yield accurate estimates of the uncertainty of predicted landmark locations, but it also yields state-of-the-art estimates for the landmark locations themselves on multiple standard face alignment datasets. Our method's estimates of the uncertainty of predicted landmark locations could be used to automatically identify input images on which face alignment fails, which can be critical for downstream tasks.

60.Deblurring using Analysis-Synthesis Networks Pair ⬇️

Blind image deblurring remains a challenging problem for modern artificial neural networks. Unlike other image restoration problems, deblurring networks fail behind the performance of existing deblurring algorithms in case of uniform and 3D blur models. This follows from the diverse and profound effect that the unknown blur-kernel has on the deblurring operator.
We propose a new architecture which breaks the deblurring network into an analysis network which estimates the blur, and a synthesis network that uses this kernel to deblur the image. Unlike existing deblurring networks, this design allows us to explicitly incorporate the blur-kernel in the network's training. In addition, we introduce new cross-correlation layers that allow better blur estimations, as well as unique components that allow the estimate blur to control the action of the synthesis deblurring action.
Evaluating the new approach over established benchmark datasets shows its ability to achieve state-of-the-art deblurring accuracy on various tests, as well as offer a major speedup in runtime.

61.Objectness-Aware One-Shot Semantic Segmentation ⬇️

While deep convolutional neural networks have led to great progress in image semantic segmentation, they typically require collecting a large number of densely-annotated images for training. Moreover, once trained, the model can only make predictions in a pre-defined set of categories. Therefore, few-shot image semantic segmentation has been explored to learn to segment from only a few annotated examples. In this paper, we tackle the challenging one-shot semantic segmentation problem by taking advantage of objectness. In order to capture prior knowledge of object and background, we first train an objectness segmentation module which generalizes well to unseen categories. Then we use the objectness module to predict the objects present in the query image, and train an objectness-aware few-shot segmentation model that takes advantage of both the object information and limited annotations of the unseen category to perform segmentation in the query image. Our method achieves a mIoU score of 57.9% and 22.6% given only one annotated example of an unseen category in PASCAL-5i and COCO-20i, outperforming related baselines overall.

62.Fingerprint Presentation Attack Detection: A Sensor and Material Agnostic Approach ⬇️

The vulnerability of automated fingerprint recognition systems to presentation attacks (PA), i.e., spoof or altered fingers, has been a growing concern, warranting the development of accurate and efficient presentation attack detection (PAD) methods. However, one major limitation of the existing PAD solutions is their poor generalization to new PA materials and fingerprint sensors, not used in training. In this study, we propose a robust PAD solution with improved cross-material and cross-sensor generalization. Specifically, we build on top of any CNN-based architecture trained for fingerprint spoof detection combined with cross-material spoof generalization using a style transfer network wrapper. We also incorporate adversarial representation learning (ARL) in deep neural networks (DNN) to learn sensor and material invariant representations for PAD. Experimental results on LivDet 2015 and 2017 public domain datasets exhibit the effectiveness of the proposed approach.

63.Efficient Scale Estimation Methods using Lightweight Deep Convolutional Neural Networks for Visual Tracking ⬇️

In recent years, visual tracking methods that are based on discriminative correlation filters (DCF) have been very promising. However, most of these methods suffer from a lack of robust scale estimation skills. Although a wide range of recent DCF-based methods exploit the features that are extracted from deep convolutional neural networks (CNNs) in their translation model, the scale of the visual target is still estimated by hand-crafted features. Whereas the exploitation of CNNs imposes a high computational burden, this paper exploits pre-trained lightweight CNNs models to propose two efficient scale estimation methods, which not only improve the visual tracking performance but also provide acceptable tracking speeds. The proposed methods are formulated based on either holistic or region representation of convolutional feature maps to efficiently integrate into DCF formulations to learn a robust scale model in the frequency domain. Moreover, against the conventional scale estimation methods with iterative feature extraction of different target regions, the proposed methods exploit proposed one-pass feature extraction processes that significantly improve the computational efficiency. Comprehensive experimental results on the OTB-50, OTB-100, TC-128 and VOT-2018 visual tracking datasets demonstrate that the proposed visual tracking methods outperform the state-of-the-art methods, effectively.

64.Beyond Background-Aware Correlation Filters: Adaptive Context Modeling by Hand-Crafted and Deep RGB Features for Visual Tracking ⬇️

In recent years, the background-aware correlation filters have achie-ved a lot of research interest in the visual target tracking. However, these methods cannot suitably model the target appearance due to the exploitation of hand-crafted features. On the other hand, the recent deep learning-based visual tracking methods have provided a competitive performance along with extensive computations. In this paper, an adaptive background-aware correlation filter-based tracker is proposed that effectively models the target appearance by using either the histogram of oriented gradients (HOG) or convolutional neural network (CNN) feature maps. The proposed method exploits the fast 2D non-maximum suppression (NMS) algorithm and the semantic information comparison to detect challenging situations. When the HOG-based response map is not reliable, or the context region has a low semantic similarity with prior regions, the proposed method constructs the CNN context model to improve the target region estimation. Furthermore, the rejection option allows the proposed method to update the CNN context model only on valid regions. Comprehensive experimental results demonstrate that the proposed adaptive method clearly outperforms the accuracy and robustness of visual target tracking compared to the state-of-the-art methods on the OTB-50, OTB-100, TC-128, UAV-123, and VOT-2015 datasets.

65.Empirical Upper Bound, Error Diagnosis and Invariance Analysis of Modern Object Detectors ⬇️

Object detection remains as one of the most notorious open problems in computer vision. Despite large strides in accuracy in recent years, modern object detectors have started to saturate on popular benchmarks raising the question of how far we can reach with deep learning tools and tricks. Here, by employing 2 state-of-the-art object detection benchmarks, and analyzing more than 15 models over 4 large scale datasets, we I) carefully determine the upper bound in AP, which is 91.6% on VOC (test2007), 78.2% on COCO (val2017), and 58.9% on OpenImages V4 (validation), regardless of the IOU threshold. These numbers are much better than the mAP of the best model (47.9% on VOC, and 46.9% on COCO; IOUs=.5:.05:.95), II) characterize the sources of errors in object detectors, in a novel and intuitive way, and find that classification error (confusion with other classes and misses) explains the largest fraction of errors and weighs more than localization and duplicate errors, and III) analyze the invariance properties of models when surrounding context of an object is removed, when an object is placed in an incongruent background, and when images are blurred or flipped vertically. We find that models generate a lot of boxes on empty regions and that context is more important for detecting small objects than larger ones. Our work taps into the tight relationship between object detection and object recognition and offers insights for building better models. Our code is publicly available at this https URL bound.git.

66.U-Net Using Stacked Dilated Convolutions for Medical Image Segmentation ⬇️

This paper proposes a novel U-Net variant using stacked dilated convolutions for medical image segmentation (SDU-Net). SDU-Net adopts the architecture of vanilla U-Net with modifications in the encoder and decoder operations (an operation indicates all the processing for feature maps of the same resolution). Unlike vanilla U-Net which incorporates two standard convolutions in each encoder/decoder operation, SDU-Net uses one standard convolution followed by multiple dilated convolutions and concatenates all dilated convolution outputs as input to the next operation. Experiments showed that SDU-Net outperformed vanilla U-Net, attention U-Net (AttU-Net), and recurrent residual U-Net (R2U-Net) in all four tested segmentation tasks while using parameters around 40% of vanilla U-Net's, 17% of AttU-Net's, and 15% of R2U-Net's.

67.Learning to Accelerate Decomposition for Multi-Directional 3D Printing ⬇️

As a strong complementary of additive manufacturing, multi-directional 3D printing has the capability of decreasing or eliminating the need for support structures. Recent work proposed a beam-guided search algorithm to find an optimized sequence of plane-clipping, which gives volume decomposition of a given 3D model. Different printing directions are employed in different regions so that a model can be fabricated with tremendously less supports (or even no support in many cases). To obtain optimized decomposition, a large beam width needs to be used in the search algorithm, which therefore leads to a very time-consuming computation. In this paper, we propose a learning framework that can accelerate the beam-guided search by using only 1/2 of the original beam width to obtain results with similar quality. Specifically, we train a classifier for each pair of candidate clipping planes based on six newly proposed feature metrics from the results of beam-guided search with large beam width. With the help of these feature metrics, both the current and the sequence-dependent information are captured by the classifier to score candidates of clipping. As a result, we can achieve around 2 times acceleration. We test and demonstrate the performance of our accelerated decomposition on a large dataset of models for 3D printing.

68.Deep Open Space Segmentation using Automotive Radar ⬇️

In this work, we propose the use of radar with advanced deep segmentation models to identify open space in parking scenarios. A publically available dataset of radar observations called SCORP was collected. Deep models are evaluated with various radar input representations. Our proposed approach achieves low memory usage and real-time processing speeds, and is thus very well suited for embedded deployment.

69.Harmony-Search and Otsu based System for Coronavirus Disease (COVID-19) Detection using Lung CT Scan Images ⬇️

Pneumonia is one of the foremost lung diseases and untreated pneumonia will lead to serious threats for all age groups. The proposed work aims to extract and evaluate the Coronavirus disease (COVID-19) caused pneumonia infection in lung using CT scans. We propose an image-assisted system to extract COVID-19 infected sections from lung CT scans (coronal view). It includes following steps: (i) Threshold filter to extract the lung region by eliminating possible artifacts; (ii) Image enhancement using Harmony-Search-Optimization and Otsu thresholding; (iii) Image segmentation to extract infected region(s); and (iv) Region-of-interest (ROI) extraction (features) from binary image to compute level of severity. The features that are extracted from ROI are then employed to identify the pixel ratio between the lung and infection sections to identify infection level of severity. The primary objective of the tool is to assist the pulmonologist not only to detect but also to help plan treatment process. As a consequence, for mass screening processing, it will help prevent diagnostic burden.

70.Automated Smartphone based System for Diagnosis of Diabetic Retinopathy ⬇️

Early diagnosis of diabetic retinopathy for treatment of the disease has been failing to reach diabetic people living in rural areas. Shortage of trained ophthalmologists, limited availability of healthcare centers, and expensiveness of diagnostic equipment are among the reasons. Although many deep learning-based automatic diagnosis of diabetic retinopathy techniques have been implemented in the literature, these methods still fail to provide a point-of-care diagnosis. This raises the need for an independent diagnostic of diabetic retinopathy that can be used by a non-expert. Recently the usage of smartphones has been increasing across the world. Automated diagnoses of diabetic retinopathy can be deployed on smartphones in order to provide an instant diagnosis to diabetic people residing in remote areas. In this paper, inception based convolutional neural network and binary decision tree-based ensemble of classifiers have been proposed and implemented to detect and classify diabetic retinopathy. The proposed method was further imported into a smartphone application for mobile-based classification, which provides an offline and automatic system for diagnosis of diabetic retinopathy.

71.Deep Learning on Chest X-ray Images to Detect and Evaluate Pneumonia Cases at the Era of COVID-19 ⬇️

Coronavirus disease 2019 (COVID-19) is an infectious disease with first symptoms similar to the flu. COVID-19 appeared first in China and very quickly spreads to the rest of the world, causing then the 2019-20 coronavirus pandemic. In many cases, this disease causes pneumonia. Since pulmonary infections can be observed through radiography images, this paper investigates deep learning methods for automatically analyzing query chest X-ray images with the hope to bring precision tools to health professionals towards screening the COVID-19 and diagnosing confirmed patients. In this context, training datasets, deep learning architectures and analysis strategies have been experimented from publicly open sets of chest X-ray images. Tailored deep learning models are proposed to detect pneumonia infection cases, notably viral cases. It is assumed that viral pneumonia cases detected during an epidemic COVID-19 context have a high probability to presume COVID-19 infections. Moreover, easy-to-apply health indicators are proposed for estimating infection status and predicting patient status from the detected pneumonia cases. Experimental results show possibilities of training deep learning models over publicly open sets of chest X-ray images towards screening viral pneumonia. Chest X-ray test images of COVID-19 infected patients are successfully diagnosed through detection models retained for their performances. The efficiency of proposed health indicators is highlighted through simulated scenarios of patients presenting infections and health problems by combining real and synthetic health data.

72.Complete CVDL Methodology for Investigating Hydrodynamic Instabilities ⬇️

In fluid dynamics, one of the most important research fields is hydrodynamic instabilities and their evolution in different flow regimes. The investigation of said instabilities is concerned with the highly non-linear dynamics. Currently, three main methods are used for understanding of such phenomenon - namely analytical models, experiments and simulations - and all of them are primarily investigated and correlated using human expertise. In this work we claim and demonstrate that a major portion of this research effort could and should be analysed using recent breakthrough advancements in the field of Computer Vision with Deep Learning (CVDL, or Deep Computer-Vision). Specifically, we target and evaluate specific state-of-the-art techniques - such as Image Retrieval, Template Matching, Parameters Regression and Spatiotemporal Prediction - for the quantitative and qualitative benefits they provide. In order to do so we focus in this research on one of the most representative instabilities, the Rayleigh-Taylor one, simulate its behaviour and create an open-sourced state-of-the-art annotated database (RayleAI). Finally, we use adjusted experimental results and novel physical loss methodologies to validate the correspondence of the predicted results to actual physical reality to prove the models efficiency. The techniques which were developed and proved in this work can be served as essential tools for physicists in the field of hydrodynamics for investigating a variety of physical systems, and also could be used via Transfer Learning to other instabilities research. A part of the techniques can be easily applied on already exist simulation results. All models as well as the data-set that was created for this work, are publicly available at: this https URL.

73.Convolutional Neural Networks based automated segmentation and labelling of the lumbar spine X-ray ⬇️

The aim of this study is to investigate the segmentation accuracies of different segmentation networks trained on 730 manually annotated lateral lumbar spine X-rays. Instance segmentation networks were compared to semantic segmentation networks. The study cohort comprised diseased spines and postoperative images with metallic implants. The average mean accuracy and mean intersection over union (IoU) was up to 3 percent better for the best performing instance segmentation model, the average pixel accuracy and weighted IoU were slightly better for the best performing semantic segmentation model. Moreover, the inferences of the instance segmentation models are easier to implement for further processing pipelines in clinical decision support.

74.Binary Neural Networks: A Survey ⬇️

The binary neural network, largely saving the storage and computation, serves as a promising technique for deploying deep models on resource-limited devices. However, the binarization inevitably causes severe information loss, and even worse, its discontinuity brings difficulty to the optimization of the deep network. To address these issues, a variety of algorithms have been proposed, and achieved satisfying progress in recent years. In this paper, we present a comprehensive survey of these algorithms, mainly categorized into the native solutions directly conducting binarization, and the optimized ones using techniques like minimizing the quantization error, improving the network loss function, and reducing the gradient error. We also investigate other practical aspects of binary neural networks such as the hardware-friendly design and the training tricks. Then, we give the evaluation and discussions on different tasks, including image classification, object detection and semantic segmentation. Finally, the challenges that may be faced in future research are prospected.

75.Two-Stage Resampling for Convolutional Neural Network Training in the Imbalanced Colorectal Cancer Image Classification ⬇️

Data imbalance remains one of the open challenges in the contemporary machine learning. It is especially prevalent in case of medical data, such as histopathological images. Traditional data-level approaches for dealing with data imbalance are ill-suited for image data: oversampling methods such as SMOTE and its derivatives lead to creation of unrealistic synthetic observations, whereas undersampling reduces the amount of available data, critical for successful training of convolutional neural networks. To alleviate the problems associated with over- and undersampling we propose a novel two-stage resampling methodology, in which we initially use the oversampling techniques in the image space to leverage a large amount of data for training of a convolutional neural network, and afterwards apply undersampling in the feature space to fine-tune the last layers of the network. Experiments conducted on a colorectal cancer image dataset indicate the usefulness of the proposed approach.

76.Teacher-Class Network: A Neural Network Compression Mechanism ⬇️

To solve the problem of the overwhelming size of Deep Neural Networks (DNN) several compression schemes have been proposed, one of them is teacher-student. Teacher-student tries to transfer knowledge from a complex teacher network to a simple student network. In this paper, we propose a novel method called a teacher-class network consisting of a single teacher and multiple student networks (i.e. class of students). Instead of transferring knowledge to one student only, the proposed method transfers a chunk of knowledge about the entire solution to each student. Our students are not trained for problem-specific logits, they are trained to mimic knowledge (dense representation) learned by the teacher network. Thus unlike the logits-based single student approach, the combined knowledge learned by the class of students can be used to solve other problems as well. These students can be designed to satisfy a given budget, e.g. for comparative purposes we kept the collective parameters of all the students less than or equivalent to that of a single student in the teacher-student approach . These small student networks are trained independently, making it possible to train and deploy models on memory deficient devices as well as on parallel processing systems such as data centers. The proposed teacher-class architecture is evaluated on several benchmark datasets including MNIST, FashionMNIST, IMDB Movie Reviews and CAMVid on multiple tasks including classification, sentiment classification and segmentation. Our approach outperforms the state-of-the-art single student approach in terms of accuracy as well as computational cost and in many cases it achieves an accuracy equivalent to the teacher network while having 10-30 times fewer parameters.

77.Autoencoders for Unsupervised Anomaly Segmentation in Brain MR Images: A Comparative Study ⬇️

Deep unsupervised representation learning has recently led to new approaches in the field of Unsupervised Anomaly Detection (UAD) in brain MRI. The main principle behind these works is to learn a model of normal anatomy by learning to compress and recover healthy data. This allows to spot abnormal structures from erroneous recoveries of compressed, potentially anomalous samples. The concept is of great interest to the medical image analysis community as it i) relieves from the need of vast amounts of manually segmented training data---a necessity for and pitfall of current supervised Deep Learning---and ii) theoretically allows to detect arbitrary, even rare pathologies which supervised approaches might fail to find. To date, the experimental design of most works hinders a valid comparison, because i) they are evaluated against different datasets and different pathologies, ii) use different image resolutions and iii) different model architectures with varying complexity. The intent of this work is to establish comparability among recent methods by utilizing a single architecture, a single resolution and the same dataset(s). Besides providing a ranking of the methods, we also try to answer questions like i) how many healthy training subjects are needed to model normality and ii) if the reviewed approaches are also sensitive to domain shift. Further, we identify open challenges and provide suggestions for future community efforts and research directions.

78.Inspector Gadget: A Data Programming-based Labeling System for Industrial Images ⬇️

As machine learning for images becomes democratized in the Software 2.0 era, one of the serious bottlenecks is securing enough labeled data for training. This problem is especially critical in a manufacturing setting where smart factories rely on machine learning for product quality control by analyzing industrial images. Such images are typically large and may only need to be partially analyzed where only a small portion is problematic (e.g., identifying defects on a surface). Since manual labeling these images is expensive, weak supervision is an attractive alternative where the idea is to generate weak labels that are not perfect, but can be produced at scale. Data programming is a recent paradigm in this category where it uses human knowledge in the form of labeling functions and combines them into a generative model. Data programming has been successful in applications based on text or structured data and can also be applied to images usually if one can find a way to convert them into structured data. In this work, we expand the horizon of data programming by directly applying it to images without this conversion, which is a common scenario for industrial applications. We propose Inspector Gadget, an image labeling system that combines crowdsourcing, data augmentation, and data programming to produce weak labels at scale for image classification. We perform experiments on real industrial image datasets and show that Inspector Gadget obtains better accuracy than state-of-the-art techniques: Snuba, GOGGLES, and self-learning baselines using convolutional neural networks (CNNs) without pre-training.

79.Iconify: Converting Photographs into Icons ⬇️

In this paper, we tackle a challenging domain conversion task between photo and icon images. Although icons often originate from real object images (i.e., photographs), severe abstractions and simplifications are applied to generate icon images by professional graphic designers. Moreover, there is no one-to-one correspondence between the two domains, for this reason we cannot use it as the ground-truth for learning a direct conversion function. Since generative adversarial networks (GAN) can undertake the problem of domain conversion without any correspondence, we test CycleGAN and UNIT to generate icons from objects segmented from photo images. Our experiments with several image datasets prove that CycleGAN learns sufficient abstraction and simplification ability to generate icon-like images.

80.Deep Attentive Generative Adversarial Network for Photo-Realistic Image De-Quantization ⬇️

Most of current display devices are with eight or higher bit-depth. However, the quality of most multimedia tools cannot achieve this bit-depth standard for the generating images. De-quantization can improve the visual quality of low bit-depth image to display on high bit-depth screen. This paper proposes DAGAN algorithm to perform super-resolution on image intensity resolution, which is orthogonal to the spatial resolution, realizing photo-realistic de-quantization via an end-to-end learning pattern. Until now, this is the first attempt to apply Generative Adversarial Network (GAN) framework for image de-quantization. Specifically, we propose the Dense Residual Self-attention (DenseResAtt) module, which is consisted of dense residual blocks armed with self-attention mechanism, to pay more attention on high-frequency information. Moreover, the series connection of sequential DenseResAtt modules forms deep attentive network with superior discriminative learning ability in image de-quantization, modeling representative feature maps to recover as much useful information as possible. In addition, due to the adversarial learning framework can reliably produce high quality natural images, the specified content loss as well as the adversarial loss are back-propagated to optimize the training of model. Above all, DAGAN is able to generate the photo-realistic high bit-depth image without banding artifacts. Experiment results on several public benchmarks prove that the DAGAN algorithm possesses ability to achieve excellent visual effect and satisfied quantitative performance.

81.Plug-and-play ISTA converges with kernel denoisers ⬇️

Plug-and-play (PnP) method is a recent paradigm for image regularization, where the proximal operator (associated with some given regularizer) in an iterative algorithm is replaced with a powerful denoiser. Algorithmically, this involves repeated inversion (of the forward model) and denoising until convergence. Remarkably, PnP regularization produces promising results for several restoration applications. However, a fundamental question in this regard is the theoretical convergence of the PnP iterations, since the algorithm is not strictly derived from an optimization framework. This question has been investigated in recent works, but there are still many unresolved problems. For example, it is not known if convergence can be guaranteed if we use generic kernel denoisers (e.g. nonlocal means) within the ISTA framework (PnP-ISTA). We prove that, under reasonable assumptions, fixed-point convergence of PnP-ISTA is indeed guaranteed for linear inverse problems such as deblurring, inpainting and superresolution (the assumptions are verifiable for inpainting). We compare our theoretical findings with existing results, validate them numerically, and explain their practical relevance.

82.Generalized Label Enhancement with Sample Correlations ⬇️

Recently, label distribution learning (LDL) has drawn much attention in machine learning, where LDL model is learned from labeled instances. Different from single-label and multi-label annotations, label distributions describe the instance by multiple labels with different intensities and accommodates to more general conditions. As most existing machine learning datasets merely provide logical labels, label distributions are unavailable in many real-world applications. To handle this problem, we propose two novel label enhancement methods, i.e., Label Enhancement with Sample Correlations (LESC) and generalized Label Enhancement with Sample Correlations (gLESC). More specifically, LESC employs a low-rank representation of samples in the feature space, and gLESC leverages a tensor multi-rank minimization to further investigate sample correlations in both the feature space and label space. Benefit from the sample correlation, the proposed method can boost the performance of LE. Extensive experiments on 14 benchmark datasets demonstrate that LESC and gLESC can achieve state-of-the-art results as compared to previous label enhancement baselines.

83.COVID-Xpert: An AI Powered Population Screening of COVID-19 Cases Using Chest Radiography Images ⬇️

With the increasing demand for millions of COVID-19 tests, Computed Tomography (CT) based test has emerged as a promising alternative to the gold standard RT-PCR test. However, it is primarily provided in emergency department and hospital settings due to the need for expensive equipment and trained radiologists. The accurate, rapid yet inexpensive test that is suitable for population screening of COVID-19 cases at mobile, urgent and primary care settings is urgently needed. Here we design a deep convolutional neural network (CNN) that extracts X-ray Chest Radiography (XCR) imaging features from large scale pneumonia and normal training cases and refine them with a small amount of COVID-19 cases to learn the imaging features that are capable of automatically discriminating COVID-19 cases from pneumonia and/or normal XCR imaging cases. We demonstrate the strong potential of our XCR based population screening approach, COVID-Xpert, for detecting COVID-19 cases through an impressive experimental performance. The trained models and information of the compiled data set are available from this https URL.

84.Dense Steerable Filter CNNs for Exploiting Rotational Symmetry in Histology Images ⬇️

Histology images are inherently symmetric under rotation, where each orientation is equally as likely to appear. However, this rotational symmetry is not widely utilised as prior knowledge in modern Convolutional Neural Networks (CNNs), resulting in data hungry models that learn independent features at each orientation. Allowing CNNs to be rotation-equivariant removes the necessity to learn this set of transformations from the data and instead frees up model capacity, allowing more discriminative features to be learned. This reduction in the number of required parameters also reduces the risk of overfitting. In this paper, we propose Dense Steerable Filter CNNs (DSF-CNNs) that use group convolutions with multiple rotated copies of each filter in a densely connected framework. Each filter is defined as a linear combination of steerable basis filters, enabling exact rotation and decreasing the number of trainable parameters compared to standard filters. We also provide the first in-depth comparison of different rotation-equivariant CNNs for histology image analysis and demonstrate the advantage of encoding rotational symmetry into modern architectures. We show that DSF-CNNs achieve state-of-the-art performance, with significantly fewer parameters, when applied to three different tasks in the area of computational pathology: breast tumour classification, colon gland segmentation and multi-tissue nuclear segmentation.

85.Evolving Normalization-Activation Layers ⬇️

Normalization layers and activation functions are critical components in deep neural networks that frequently co-locate with each other. Instead of designing them separately, we unify them into a single computation graph, and evolve its structure starting from low-level primitives. Our layer search algorithm leads to the discovery of EvoNorms, a set of new normalization-activation layers that go beyond existing design patterns. Several of these layers enjoy the property of being independent from the batch statistics. Our experiments show that EvoNorms not only excel on a variety of image classification models including ResNets, MobileNets and EfficientNets, but also transfer well to Mask R-CNN for instance segmentation and BigGAN for image synthesis, outperforming BatchNorm and GroupNorm based layers by a significant margin in many cases.