1.Face, Body, Voice: Video Person-Clustering with Multiple Modalities ⬇️
The objective of this work is person-clustering in videos -- grouping characters according to their identity. Previous methods focus on the narrower task of face-clustering, and for the most part ignore other cues such as the person's voice, their overall appearance (hair, clothes, posture), and the editing structure of the videos. Similarly, most current datasets evaluate only the task of face-clustering, rather than person-clustering. This limits their applicability to downstream applications such as story understanding which require person-level, rather than only face-level, reasoning. In this paper we make contributions to address both these deficiencies: first, we introduce a Multi-Modal High-Precision Clustering algorithm for person-clustering in videos using cues from several modalities (face, body, and voice). Second, we introduce a Video Person-Clustering dataset, for evaluating multi-modal person-clustering. It contains body-tracks for each annotated character, face-tracks when visible, and voice-tracks when speaking, with their associated features. The dataset is by far the largest of its kind, and covers films and TV-shows representing a wide range of demographics. Finally, we show the effectiveness of using multiple modalities for person-clustering, explore the use of this new broad task for story understanding through character co-occurrences, and achieve a new state of the art on all available datasets for face and person-clustering.
2.AnaXNet: Anatomy Aware Multi-label Finding Classification in Chest X-ray ⬇️
Radiologists usually observe anatomical regions of chest X-ray images as well as the overall image before making a decision. However, most existing deep learning models only look at the entire X-ray image for classification, failing to utilize important anatomical information. In this paper, we propose a novel multi-label chest X-ray classification model that accurately classifies the image finding and also localizes the findings to their correct anatomical regions. Specifically, our model consists of two modules, the detection module and the anatomical dependency module. The latter utilizes graph convolutional networks, which enable our model to learn not only the label dependency but also the relationship between the anatomical regions in the chest X-ray. We further utilize a method to efficiently create an adjacency matrix for the anatomical regions using the correlation of the label across the different regions. Detailed experiments and analysis of our results show the effectiveness of our method when compared to the current state-of-the-art multi-label chest X-ray image classification methods while also providing accurate location information.
3.BodyPressure -- Inferring Body Pose and Contact Pressure from a Depth Image ⬇️
Contact pressure between the human body and its surroundings has important implications. For example, it plays a role in comfort, safety, posture, and health. We present a method that infers contact pressure between a human body and a mattress from a depth image. Specifically, we focus on using a depth image from a downward facing camera to infer pressure on a body at rest in bed occluded by bedding, which is directly applicable to the prevention of pressure injuries in healthcare. Our approach involves augmenting a real dataset with synthetic data generated via a soft-body physics simulation of a human body, a mattress, a pressure sensing mat, and a blanket. We introduce a novel deep network that we trained on an augmented dataset and evaluated with real data. The network contains an embedded human body mesh model and uses a white-box model of depth and pressure image generation. Our network successfully infers body pose, outperforming prior work. It also infers contact pressure across a 3D mesh model of the human body, which is a novel capability, and does so in the presence of occlusion from blankets.
4.Probing the Effect of Selection Bias on NN Generalization with a Thought Experiment ⬇️
Learned networks in the domain of visual recognition and cognition impress in part because even though they are trained with datasets many orders of magnitude smaller than the full population of possible images, they exhibit sufficient generalization to be applicable to new and previously unseen data. Although many have examined issues regarding generalization from several perspectives, we wondered If a network is trained with a biased dataset that misses particular samples corresponding to some defining domain attribute, can it generalize to the full domain from which that training dataset was extracted? It is certainly true that in vision, no current training set fully captures all visual information and this may lead to Selection Bias. Here, we try a novel approach in the tradition of the Thought Experiment. We run this thought experiment on a real domain of visual objects that we can fully characterize and look at specific gaps in training data and their impact on performance requirements. Our thought experiment points to three conclusions: first, that generalization behavior is dependent on how sufficiently the particular dimensions of the domain are represented during training; second, that the utility of any generalization is completely dependent on the acceptable system error; and third, that specific visual features of objects, such as pose orientations out of the imaging plane or colours, may not be recoverable if not represented sufficiently in a training set. Any currently observed generalization in modern deep learning networks may be more the result of coincidental alignments and whose utility needs to be confirmed with respect to a system's performance specification. Our Thought Experiment Probe approach, coupled with the resulting Bias Breakdown can be very informative towards understanding the impact of biases.
5.PLSM: A Parallelized Liquid State Machine for Unintentional Action Detection ⬇️
Reservoir Computing (RC) offers a viable option to deploy AI algorithms on low-end embedded system platforms. Liquid State Machine (LSM) is a bio-inspired RC model that mimics the cortical microcircuits and uses spiking neural networks (SNN) that can be directly realized on neuromorphic hardware. In this paper, we present a novel Parallelized LSM (PLSM) architecture that incorporates spatio-temporal read-out layer and semantic constraints on model output. To the best of our knowledge, such a formulation has been done for the first time in literature, and it offers a computationally lighter alternative to traditional deep-learning models. Additionally, we also present a comprehensive algorithm for the implementation of parallelizable SNNs and LSMs that are GPU-compatible. We implement the PLSM model to classify unintentional/accidental video clips, using the Oops dataset. From the experimental results on detecting unintentional action in video, it can be observed that our proposed model outperforms a self-supervised model and a fully supervised traditional deep learning model. All the implemented codes can be found at our repository this https URL.
6.Classification of Urban Morphology with Deep Learning: Application on Urban Vitality ⬇️
There is a prevailing trend to study urban morphology quantitatively thanks to the growing accessibility to various forms of spatial big data, increasing computing power, and use cases benefiting from such information. The methods developed up to now measure urban morphology with numerical indices describing density, proportion, and mixture, but they do not directly represent morphological features from human's visual and intuitive perspective. We take the first step to bridge the gap by proposing a deep learning-based technique to automatically classify road networks into four classes on a visual basis. The method is implemented by generating an image of the street network (Colored Road Hierarchy Diagram), which we introduce in this paper, and classifying it using a deep convolutional neural network (ResNet-34). The model achieves an overall classification accuracy of 0.875. Nine cities around the world are selected as the study areas and their road networks are acquired from OpenStreetMap. Latent subgroups among the cities are uncovered through a clustering on the percentage of each road network category. In the subsequent part of the paper, we focus on the usability of such classification: the effectiveness of our human perception augmentation is examined by a case study of urban vitality prediction. An advanced tree-based regression model is for the first time designated to establish the relationship between morphological indices and vitality indicators. A positive effect of human perception augmentation is detected in the comparative experiment of baseline model and augmented model. This work expands the toolkit of quantitative urban morphology study with new techniques, supporting further studies in the future.
7.Joint Face Image Restoration and Frontalization for Recognition ⬇️
In real-world scenarios, many factors may harm face recognition performance, e.g., large pose, bad illumination,low resolution, blur and noise. To address these challenges, previous efforts usually first restore the low-quality faces to high-quality ones and then perform face recognition. However, most of these methods are stage-wise, which is sub-optimal and deviates from the reality. In this paper, we address all these challenges jointly for unconstrained face recognition. We propose an Multi-Degradation Face Restoration (MDFR) model to restore frontalized high-quality faces from the given low-quality ones under arbitrary facial poses, with three distinct novelties. First, MDFR is a well-designed encoder-decoder architecture which extracts feature representation from an input face image with arbitrary low-quality factors and restores it to a high-quality counterpart. Second, MDFR introduces a pose residual learning strategy along with a 3D-based Pose Normalization Module (PNM), which can perceive the pose gap between the input initial pose and its real-frontal pose to guide the face frontalization. Finally, MDFR can generate frontalized high-quality face images by a single unified network, showing a strong capability of preserving face identity. Qualitative and quantitative experiments on both controlled and in-the-wild benchmarks demonstrate the superiority of MDFR over state-of-the-art methods on both face frontalization and face restoration.
8.Empirical Analysis of Image Caption Generation using Deep Learning ⬇️
Automated image captioning is one of the applications of Deep Learning which involves fusion of work done in computer vision and natural language processing, and it is typically performed using Encoder-Decoder architectures. In this project, we have implemented and experimented with various flavors of multi-modal image captioning networks where ResNet101, DenseNet121 and VGG19 based CNN Encoders and Attention based LSTM Decoders were explored. We have studied the effect of beam size and the use of pretrained word embeddings and compared them to baseline CNN encoder and RNN decoder architecture. The goal is to analyze the performance of each approach using various evaluation metrics including BLEU, CIDEr, ROUGE and METEOR. We have also explored model explainability using Visual Attention Maps (VAM) to highlight parts of the images which has maximum contribution for predicting each word of the generated caption.
9.Multi-Perspective Anomaly Detection ⬇️
Multi-view classification is inspired by the behavior of humans, especially when fine-grained features or in our case rarely occurring anomalies are to be detected. Current contributions point to the problem of how high-dimensional data can be fused. In this work, we build upon the deep support vector data description algorithm and address multi-perspective anomaly detection using three different fusion techniques i.e. early fusion, late fusion, and late fusion with multiple decoders. We employ different augmentation techniques with a denoising process to deal with scarce one-class data, which further improves the performance (ROC AUC = 80%). Furthermore, we introduce the dices dataset that consists of over 2000 grayscale images of falling dices from multiple perspectives, with 5% of the images containing rare anomalies (e.g. drill holes, sawing, or scratches). We evaluate our approach on the new dices dataset using images from two different perspectives and also benchmark on the standard MNIST dataset. Extensive experiments demonstrate that our proposed approach exceeds the state-of-the-art on both the MNIST and dices datasets. To the best of our knowledge, this is the first work that focuses on addressing multi-perspective anomaly detection in images by jointly using different perspectives together with one single objective function for anomaly detection.
10.DeepAVO: Efficient Pose Refining with Feature Distilling for Deep Visual Odometry ⬇️
The technology for Visual Odometry (VO) that estimates the position and orientation of the moving object through analyzing the image sequences captured by on-board cameras, has been well investigated with the rising interest in autonomous driving. This paper studies monocular VO from the perspective of Deep Learning (DL). Unlike most current learning-based methods, our approach, called DeepAVO, is established on the intuition that features contribute discriminately to different motion patterns. Specifically, we present a novel four-branch network to learn the rotation and translation by leveraging Convolutional Neural Networks (CNNs) to focus on different quadrants of optical flow input. To enhance the ability of feature selection, we further introduce an effective channel-spatial attention mechanism to force each branch to explicitly distill related information for specific Frame to Frame (F2F) motion estimation. Experiments on various datasets involving outdoor driving and indoor walking scenarios show that the proposed DeepAVO outperforms the state-of-the-art monocular methods by a large margin, demonstrating competitive performance to the stereo VO algorithm and verifying promising potential for generalization.
11.DeepDarts: Modeling Keypoints as Objects for Automatic Scorekeeping in Darts using a Single Camera ⬇️
Existing multi-camera solutions for automatic scorekeeping in steel-tip darts are very expensive and thus inaccessible to most players. Motivated to develop a more accessible low-cost solution, we present a new approach to keypoint detection and apply it to predict dart scores from a single image taken from any camera angle. This problem involves detecting multiple keypoints that may be of the same class and positioned in close proximity to one another. The widely adopted framework for regressing keypoints using heatmaps is not well-suited for this task. To address this issue, we instead propose to model keypoints as objects. We develop a deep convolutional neural network around this idea and use it to predict dart locations and dartboard calibration points within an overall pipeline for automatic dart scoring, which we call DeepDarts. Additionally, we propose several task-specific data augmentation strategies to improve the generalization of our method. As a proof of concept, two datasets comprising 16k images originating from two different dartboard setups were manually collected and annotated to evaluate the system. In the primary dataset containing 15k images captured from a face-on view of the dartboard using a smartphone, DeepDarts predicted the total score correctly in 94.7% of the test images. In a second more challenging dataset containing limited training data (830 images) and various camera angles, we utilize transfer learning and extensive data augmentation to achieve a test accuracy of 84.0%. Because DeepDarts relies only on single images, it has the potential to be deployed on edge devices, giving anyone with a smartphone access to an automatic dart scoring system for steel-tip darts. The code and datasets are available.
12.Flexible Compositional Learning of Structured Visual Concepts ⬇️
Humans are highly efficient learners, with the ability to grasp the meaning of a new concept from just a few examples. Unlike popular computer vision systems, humans can flexibly leverage the compositional structure of the visual world, understanding new concepts as combinations of existing concepts. In the current paper, we study how people learn different types of visual compositions, using abstract visual forms with rich relational structure. We find that people can make meaningful compositional generalizations from just a few examples in a variety of scenarios, and we develop a Bayesian program induction model that provides a close fit to the behavioral data. Unlike past work examining special cases of compositionality, our work shows how a single computational approach can account for many distinct types of compositional generalization.
13.M4Depth: A motion-based approach for monocular depth estimation on video sequences ⬇️
Getting the distance to objects is crucial for autonomous vehicles. In instances where depth sensors cannot be used, this distance has to be estimated from RGB cameras. As opposed to cars, the task of estimating depth from on-board mounted cameras is made complex on drones because of the lack of constrains on motion during flights. %In the case of drones, this task is even more complex than for car-mounted cameras since the camera motion is unconstrained. In this paper, we present a method to estimate the distance of objects seen by an on-board mounted camera by using its RGB video stream and drone motion information. Our method is built upon a pyramidal convolutional neural network architecture and uses time recurrence in pair with geometric constraints imposed by motion to produce pixel-wise depth maps. %from a RGB video stream of a camera attached to the drone In our architecture, each level of the pyramid is designed to produce its own depth estimate based on past observations and information provided by the previous level in the pyramid. We introduce a spatial reprojection layer to maintain the spatio-temporal consistency of the data between the levels. We analyse the performance of our approach on Mid-Air, a public drone dataset featuring synthetic drone trajectories recorded in a wide variety of unstructured outdoor environments. Our experiments show that our network outperforms state-of-the-art depth estimation methods and that the use of motion information is the main contributing factor for this improvement. The code of our method is publicly available on GitHub; see $\href{this https URL}{\text{this https URL}}$
14.Biologically Inspired Semantic Lateral Connectivity for Convolutional Neural Networks ⬇️
Lateral connections play an important role for sensory processing in visual cortex by supporting discriminable neuronal responses even to highly similar features. In the present work, we show that establishing a biologically inspired Mexican hat lateral connectivity profile along the filter domain can significantly improve the classification accuracy of a variety of lightweight convolutional neural networks without the addition of trainable network parameters. Moreover, we demonstrate that it is possible to analytically determine the stationary distribution of modulated filter activations and thereby avoid using recurrence for modeling temporal dynamics. We furthermore reveal that the Mexican hat connectivity profile has the effect of ordering filters in a sequence resembling the topographic organization of feature selectivity in early visual cortex. In an ordered filter sequence, this profile then sharpens the filters' tuning curves.
15.Weakly-Supervised Physically Unconstrained Gaze Estimation ⬇️
A major challenge for physically unconstrained gaze estimation is acquiring training data with 3D gaze annotations for in-the-wild and outdoor scenarios. In contrast, videos of human interactions in unconstrained environments are abundantly available and can be much more easily annotated with frame-level activity labels. In this work, we tackle the previously unexplored problem of weakly-supervised gaze estimation from videos of human interactions. We leverage the insight that strong gaze-related geometric constraints exist when people perform the activity of "looking at each other" (LAEO). To acquire viable 3D gaze supervision from LAEO labels, we propose a training algorithm along with several novel loss functions especially designed for the task. With weak supervision from two large scale CMU-Panoptic and AVA-LAEO activity datasets, we show significant improvements in (a) the accuracy of semi-supervised gaze estimation and (b) cross-domain generalization on the state-of-the-art physically unconstrained in-the-wild Gaze360 gaze estimation benchmark. We open source our code at this https URL.
16.A Spatio-temporal Attention-based Model for Infant Movement Assessment from Videos ⬇️
The absence or abnormality of fidgety movements of joints or limbs is strongly indicative of cerebral palsy in infants. Developing computer-based methods for assessing infant movements in videos is pivotal for improved cerebral palsy screening. Most existing methods use appearance-based features and are thus sensitive to strong but irrelevant signals caused by background clutter or a moving camera. Moreover, these features are computed over the whole frame, thus they measure gross whole body movements rather than specific joint/limb motion.
Addressing these challenges, we develop and validate a new method for fidgety movement assessment from consumer-grade videos using human poses extracted from short clips. Human poses capture only relevant motion profiles of joints and limbs and are thus free from irrelevant appearance artifacts. The dynamics and coordination between joints are modeled using spatio-temporal graph convolutional networks. Frames and body parts that contain discriminative information about fidgety movements are selected through a spatio-temporal attention mechanism. We validate the proposed model on the cerebral palsy screening task using a real-life consumer-grade video dataset collected at an Australian hospital through the Cerebral Palsy Alliance, Australia. Our experiments show that the proposed method achieves the ROC-AUC score of 81.87%, significantly outperforming existing competing methods with better interpretability.
17.Semi-supervised, Topology-Aware Segmentation of Tubular Structures from Live Imaging 3D Microscopy ⬇️
Motivated by a challenging tubular network segmentation task, this paper tackles two commonly encountered problems in biomedical imaging: Topological consistency of the segmentation, and limited annotations. We propose a topological score which measures both topological and geometric consistency between the predicted and ground truth segmentations, applied for model selection and validation. We apply our topological score in three scenarios: i. a U-net ii. a U-net pretrained on an autoencoder, and iii. a semisupervised U-net architecture, which offers a straightforward approach to jointly training the network both as an autoencoder and a segmentation algorithm. This allows us to utilize un-annotated data for training a representation that generalizes across test data variability, in spite of our annotated training data having very limited variation. Our contributions are validated on a challenging segmentation task, locating tubular structures in the fetal pancreas from noisy live imaging confocal microscopy.
18.An Attractor-Guided Neural Networks for Skeleton-Based Human Motion Prediction ⬇️
Joint relation modeling is a curial component in human motion prediction. Most existing methods tend to design skeletal-based graphs to build the relations among joints, where local interactions between joint pairs are well learned. However, the global coordination of all joints, which reflects human motion's balance property, is usually weakened because it is learned from part to whole progressively and asynchronously. Thus, the final predicted motions are sometimes unnatural. To tackle this issue, we learn a medium, called balance attractor (BA), from the spatiotemporal features of motion to characterize the global motion features, which is subsequently used to build new joint relations. Through the BA, all joints are related synchronously, and thus the global coordination of all joints can be better learned. Based on the BA, we propose our framework, referred to Attractor-Guided Neural Network, mainly including Attractor-Based Joint Relation Extractor (AJRE) and Multi-timescale Dynamics Extractor (MTDE). The AJRE mainly includes Global Coordination Extractor (GCE) and Local Interaction Extractor (LIE). The former presents the global coordination of all joints, and the latter encodes local interactions between joint pairs. The MTDE is designed to extract dynamic information from raw position information for effective prediction. Extensive experiments show that the proposed framework outperforms state-of-the-art methods in both short and long-term predictions in H3.6M, CMU-Mocap, and 3DPW.
19.An Empirical Study of Vehicle Re-Identification on the AI City Challenge ⬇️
This paper introduces our solution for the Track2 in AI City Challenge 2021 (AICITY21). The Track2 is a vehicle re-identification (ReID) task with both the real-world data and synthetic data. We mainly focus on four points, i.e. training data, unsupervised domain-adaptive (UDA) training, post-processing, model ensembling in this challenge. (1) Both cropping training data and using synthetic data can help the model learn more discriminative features. (2) Since there is a new scenario in the test set that dose not appear in the training set, UDA methods perform well in the challenge. (3) Post-processing techniques including re-ranking, image-to-track retrieval, inter-camera fusion, etc, significantly improve final performance. (4) We ensemble CNN-based models and transformer-based models which provide different representation diversity. With aforementioned techniques, our method finally achieves 0.7445 mAP score, yielding the first place in the competition. Codes are available at this https URL.
20.Simple Transparent Adversarial Examples ⬇️
There has been a rise in the use of Machine Learning as a Service (MLaaS) Vision APIs as they offer multiple services including pre-built models and algorithms, which otherwise take a huge amount of resources if built from scratch. As these APIs get deployed for high-stakes applications, it's very important that they are robust to different manipulations. Recent works have only focused on typical adversarial attacks when evaluating the robustness of vision APIs. We propose two new aspects of adversarial image generation methods and evaluate them on the robustness of Google Cloud Vision API's optical character recognition service and object detection APIs deployed in real-world settings such as this http URL, this http URL, Google Cloud Vision API, and Microsoft Azure's Computer Vision API. Specifically, we go beyond the conventional small-noise adversarial attacks and introduce secret embedding and transparent adversarial examples as a simpler way to evaluate robustness. These methods are so straightforward that even non-specialists can craft such attacks. As a result, they pose a serious threat where APIs are used for high-stakes applications. Our transparent adversarial examples successfully evade state-of-the art object detections APIs such as Azure Cloud Vision (attack success rate 52%) and Google Cloud Vision (attack success rate 36%). 90% of the images have a secret embedded text that successfully fools the vision of time-limited humans but is detected by Google Cloud Vision API's optical character recognition. Complementing to current research, our results provide simple but unconventional methods on robustness evaluation.
21.Crowd Counting by Self-supervised Transfer Colorization Learning and Global Prior Classification ⬇️
Labeled crowd scene images are expensive and scarce. To significantly reduce the requirement of the labeled images, we propose ColorCount, a novel CNN-based approach by combining self-supervised transfer colorization learning and global prior classification to leverage the abundantly available unlabeled data. The self-supervised colorization branch learns the semantics and surface texture of the image by using its color components as pseudo labels. The classification branch extracts global group priors by learning correlations among image clusters. Their fused resultant discriminative features (global priors, semantics and textures) provide ample priors for counting, hence significantly reducing the requirement of labeled images. We conduct extensive experiments on four challenging benchmarks. ColorCount achieves much better performance as compared with other unsupervised approaches. Its performance is close to the supervised baseline with substantially less labeled data (10% of the original one).
22.A Connected Component Labelling algorithm for multi-pixel per clock cycle video strea ⬇️
This work describes the hardware implementation of a connected component labelling (CCL) module in reprogammable logic. The main novelty of the design is the "full", i.e. without any simplifications, support of a 4 pixel per clock format (4 ppc) and real-time processing of a 4K/UltraHD video stream (3840 x 2160 pixels) at 60 frames per second. To achieve this, a special labelling method was designed and a functionality that stops the input data stream in order to process pixel groups which require writing more than one merger into the equivalence table. The proposed module was verified in simulation and in hardware on the Xilinx Zynq Ultrascale+ MPSoC chip on the ZCU104 evaluation board.
23.Content-adaptive Representation Learning for Fast Image Super-resolution ⬇️
Deep convolutional networks have attracted great attention in image restoration and enhancement. Generally, restoration quality has been improved by building more and more convolutional block. However, these methods mostly learn a specific model to handle all images and ignore difficulty diversity. In other words, an area in the image with high frequency tend to lose more information during compressing while an area with low frequency tends to lose less. In this article, we adrress the efficiency issue in image SR by incorporating a patch-wise rolling network(PRN) to content-adaptively recover images according to difficulty levels. In contrast to existing studies that ignore difficulty diversity, we adopt different stage of a neural network to perform image restoration. In addition, we propose a rolling strategy that utilizes the parameters of each stage more flexible. Extensive experiments demonstrate that our model not only shows a significant acceleration but also maintain state-of-the-art performance.
24.More Than Just Attention: Learning Cross-Modal Attentions with Contrastive Constraints ⬇️
Attention mechanisms have been widely applied to cross-modal tasks such as image captioning and information retrieval, and have achieved remarkable improvements due to its capability to learn fine-grained relevance across different modalities. However, existing attention models could be sub-optimal and lack preciseness because there is no direct supervision involved during training. In this work, we propose Contrastive Content Re-sourcing (CCR) and Contrastive Content Swapping (CCS) constraints to address such limitation. These constraints supervise the training of attention models in a contrastive learning manner without requiring explicit attention annotations. Additionally, we introduce three metrics, namely Attention Precision, Recall and F1-Score, to quantitatively evaluate the attention quality. We evaluate the proposed constraints with cross-modal retrieval (image-text matching) task. The experiments on both Flickr30k and MS-COCO datasets demonstrate that integrating these attention constraints into two state-of-the-art attention-based models improves the model performance in terms of both retrieval accuracy and attention metrics.
25.AGSFCOS: Based on attention mechanism and Scale-Equalizing pyramid network of object detection ⬇️
Recently, the anchor-free object detection model has shown great potential for accuracy and speed to exceed anchor-based object detection. Therefore, two issues are mainly studied in this article: (1) How to let the backbone network in the anchor-free object detection model learn feature extraction? (2) How to make better use of the feature pyramid network? In order to solve the above problems, Experiments show that our model has a certain improvement in accuracy compared with the current popular detection models on the COCO dataset, the designed attention mechanism module can capture contextual information well, improve detection accuracy, and use sepc network to help balance abstract and detailed information, and reduce the problem of semantic gap in the feature pyramid network. Whether it is anchor-based network model YOLOv3, Faster RCNN, or anchor-free network model Foveabox, FSAF, FCOS. Our optimal model can get 39.5% COCO AP under the background of ResNet50.
26.Intra-Model Collaborative Learning of Neural Networks ⬇️
Recently, collaborative learning proposed by Song and Chai has achieved remarkable improvements in image classification tasks by simultaneously training multiple classifier heads. However, huge memory footprints required by such multi-head structures may hinder the training of large-capacity baseline models. The natural question is how to achieve collaborative learning within a single network without duplicating any modules. In this paper, we propose four ways of collaborative learning among different parts of a single network with negligible engineering efforts. To improve the robustness of the network, we leverage the consistency of the output layer and intermediate layers for training under the collaborative learning framework. Besides, the similarity of intermediate representation and convolution kernel is also introduced to reduce the reduce redundant in a neural network. Compared to the method of Song and Chai, our framework further considers the collaboration inside a single model and takes smaller overhead. Extensive experiments on Cifar-10, Cifar-100, ImageNet32 and STL-10 corroborate the effectiveness of these four ways separately while combining them leads to further improvements. In particular, test errors on the STL-10 dataset are decreased by
$9.28%$ and$5.45%$ for ResNet-18 and VGG-16 respectively. Moreover, our method is proven to be robust to label noise with experiments on Cifar-10 dataset. For example, our method has$3.53%$ higher performance under$50%$ noise ratio setting.
27.A low-rank representation for unsupervised registration of medical images ⬇️
Registration networks have shown great application potentials in medical image analysis. However, supervised training methods have a great demand for large and high-quality labeled datasets, which is time-consuming and sometimes impractical due to data sharing issues. Unsupervised image registration algorithms commonly employ intensity-based similarity measures as loss functions without any manual annotations. These methods estimate the parameterized transformations between pairs of moving and fixed images through the optimization of the network parameters during training. However, these methods become less effective when the image quality varies, e.g., some images are corrupted by substantial noise or artifacts. In this work, we propose a novel approach based on a low-rank representation, i.e., Regnet-LRR, to tackle the problem. We project noisy images into a noise-free low-rank space, and then compute the similarity between the images. Based on the low-rank similarity measure, we train the registration network to predict the dense deformation fields of noisy image pairs. We highlight that the low-rank projection is reformulated in a way that the registration network can successfully update gradients. With two tasks, i.e., cardiac and abdominal intra-modality registration, we demonstrate that the low-rank representation can boost the generalization ability and robustness of models as well as bring significant improvements in noisy data registration scenarios.
28.Egocentric Activity Recognition and Localization on a 3D Map ⬇️
Given a video captured from a first person perspective and recorded in a familiar environment, can we recognize what the person is doing and identify where the action occurs in the 3D space? We address this challenging problem of jointly recognizing and localizing actions of a mobile user on a known 3D map from egocentric videos. To this end, we propose a novel deep probabilistic model. Our model takes the inputs of a Hierarchical Volumetric Representation (HVR) of the environment and an egocentric video, infers the 3D action location as a latent variable, and recognizes the action based on the video and contextual cues surrounding its potential locations. To evaluate our model, we conduct extensive experiments on a newly collected egocentric video dataset, in which both human naturalistic actions and photo-realistic 3D environment reconstructions are captured. Our method demonstrates strong results on both action recognition and 3D action localization across seen and unseen environments. We believe our work points to an exciting research direction in the intersection of egocentric vision, and 3D scene understanding.
29.DeepCAD: A Deep Generative Network for Computer-Aided Design Models ⬇️
Deep generative models of 3D shapes have received a great deal of research interest. Yet, almost all of them generate discrete shape representations, such as voxels, point clouds, and polygon meshes. We present the first 3D generative model for a drastically different shape representation -- describing a shape as a sequence of computer-aided design (CAD) operations. Unlike meshes and point clouds, CAD models encode the user creation process of 3D shapes, widely used in numerous industrial and engineering design tasks. However, the sequential and irregular structure of CAD operations poses significant challenges for existing 3D generative models. Drawing an analogy between CAD operations and natural language, we propose a CAD generative network based on the Transformer. We demonstrate the performance of our model for both shape autoencoding and random shape generation. To train our network, we create a new CAD dataset consisting of 179,133 models and their CAD construction sequences. We have made this dataset publicly available to promote future research on this topic.
30.Generalized Few-Shot Object Detection without Forgetting ⬇️
Recently few-shot object detection is widely adopted to deal with data-limited situations. While most previous works merely focus on the performance on few-shot categories, we claim that detecting all classes is crucial as test samples may contain any instances in realistic applications, which requires the few-shot detector to learn new concepts without forgetting. Through analysis on transfer learning based methods, some neglected but beneficial properties are utilized to design a simple yet effective few-shot detector, Retentive R-CNN. It consists of Bias-Balanced RPN to debias the pretrained RPN and Re-detector to find few-shot class objects without forgetting previous knowledge. Extensive experiments on few-shot detection benchmarks show that Retentive R-CNN significantly outperforms state-of-the-art methods on overall performance among all settings as it can achieve competitive results on few-shot classes and does not degrade the base class performance at all. Our approach has demonstrated that the long desired never-forgetting learner is available in object detection.
31.Content-Augmented Feature Pyramid Network with Light Linear Transformers ⬇️
Recently, plenty of work has tried to introduce transformers into computer vision tasks, with good results. Unlike classic convolution networks, which extract features within a local receptive field, transformers can adaptively aggregate similar features from a global view using self-attention mechanism. For object detection, Feature Pyramid Network (FPN) proposes feature interaction across layers and proves its extremely importance. However, its interaction is still in a local manner, which leaves a lot of room for improvement. Since transformer was originally designed for NLP tasks, adapting processing subject directly from text to image will cause unaffordable computation and space overhead. In this paper, we utilize a linearized attention function to overcome above problems and build a novel architecture, named Content-Augmented Feature Pyramid Network (CA-FPN), which proposes a global content extraction module and deeply combines with FPN through light linear transformers. What's more, light transformers can further make the application of multi-head attention mechanism easier. Most importantly, our CA-FPN can be readily plugged into existing FPN-based models. Extensive experiments on the challenging COCO object detection dataset demonstrated that our CA-FPN significantly outperforms competitive baselines without bells and whistles. Code will be made publicly available.
32.Anabranch Network for Camouflaged Object Segmentation ⬇️
Camouflaged objects attempt to conceal their texture into the background and discriminating them from the background is hard even for human beings. The main objective of this paper is to explore the camouflaged object segmentation problem, namely, segmenting the camouflaged object(s) for a given image. This problem has not been well studied in spite of a wide range of potential applications including the preservation of wild animals and the discovery of new species, surveillance systems, search-and-rescue missions in the event of natural disasters such as earthquakes, floods or hurricanes. This paper addresses a new challenging problem of camouflaged object segmentation. To address this problem, we provide a new image dataset of camouflaged objects for benchmarking purposes. In addition, we propose a general end-to-end network, called the Anabranch Network, that leverages both classification and segmentation tasks. Different from existing networks for segmentation, our proposed network possesses the second branch for classification to predict the probability of containing camouflaged object(s) in an image, which is then fused into the main branch for segmentation to boost up the segmentation accuracy. Extensive experiments conducted on the newly built dataset demonstrate the effectiveness of our network using various fully convolutional networks. \url{this https URL}
33.Superpixel-based Domain-Knowledge Infusion in Computer Vision ⬇️
Superpixels are higher-order perceptual groups of pixels in an image, often carrying much more information than raw pixels. There is an inherent relational structure to the relationship among different superpixels of an image. This relational information can convey some form of domain information about the image, e.g. relationship between superpixels representing two eyes in a cat image. Our interest in this paper is to construct computer vision models, specifically those based on Deep Neural Networks (DNNs) to incorporate these superpixels information. We propose a methodology to construct a hybrid model that leverages (a) Convolutional Neural Network (CNN) to deal with spatial information in an image, and (b) Graph Neural Network (GNN) to deal with relational superpixel information in the image. The proposed deep model is learned using a generic hybrid loss function that we call a `hybrid' loss. We evaluate the predictive performance of our proposed hybrid vision model on four popular image classification datasets: MNIST, FMNIST, CIFAR-10 and CIFAR-100. Moreover, we evaluate our method on three real-world classification tasks: COVID-19 X-Ray Detection, LFW Face Recognition, and SOCOFing Fingerprint Identification. The results demonstrate that the relational superpixel information provided via a GNN could improve the performance of standard CNN-based vision systems.
34.VTNet: Visual Transformer Network for Object Goal Navigation ⬇️
Object goal navigation aims to steer an agent towards a target object based on observations of the agent. It is of pivotal importance to design effective visual representations of the observed scene in determining navigation actions. In this paper, we introduce a Visual Transformer Network (VTNet) for learning informative visual representation in navigation. VTNet is a highly effective structure that embodies two key properties for visual representations: First, the relationships among all the object instances in a scene are exploited; Second, the spatial locations of objects and image regions are emphasized so that directional navigation signals can be learned. Furthermore, we also develop a pre-training scheme to associate the visual representations with navigation signals, and thus facilitate navigation policy learning. In a nutshell, VTNet embeds object and region features with their location cues as spatial-aware descriptors and then incorporates all the encoded descriptors through attention operations to achieve informative representation for navigation. Given such visual representations, agents are able to explore the correlations between visual observations and navigation actions. For example, an agent would prioritize "turning right" over "turning left" when the visual representation emphasizes on the right side of activation map. Experiments in the artificial environment AI2-Thor demonstrate that VTNet significantly outperforms state-of-the-art methods in unseen testing environments.
35.End-to-End Unsupervised Document Image Blind Denoising ⬇️
Removing noise from scanned pages is a vital step before their submission to optical character recognition (OCR) system. Most available image denoising methods are supervised where the pairs of noisy/clean pages are required. However, this assumption is rarely met in real settings. Besides, there is no single model that can remove various noise types from documents. Here, we propose a unified end-to-end unsupervised deep learning model, for the first time, that can effectively remove multiple types of noise, including salt & pepper noise, blurred and/or faded text, as well as watermarks from documents at various levels of intensity. We demonstrate that the proposed model significantly improves the quality of scanned images and the OCR of the pages on several test datasets.
36.Unsupervised learning of text line segmentationby differentiating coarse patterns ⬇️
Despite recent advances in the field of supervised deep learning for text line segmentation, unsupervised deep learning solutions are beginning to gain popularity. In this paper, we present an unsupervised deep learning method that embeds document image patches to a compact Euclidean space where distances correspond to a coarse text line pattern similarity. Once this space has been produced, text line segmentation can be easily implemented using standard techniques with the embedded feature vectors. To train the model, we extract random pairs of document image patches with the assumption that neighbour patches contain a similar coarse trend of text lines, whereas if one of them is rotated, they contain different coarse trends of text lines. Doing well on this task requires the model to learn to recognize the text lines and their salient parts. The benefit of our approach is zero manual labelling effort. We evaluate the method qualitatively and quantitatively on several variants of text line segmentation datasets to demonstrate its effectivity.
37.Birds of a Feather: Capturing Avian Shape Models from Images ⬇️
Animals are diverse in shape, but building a deformable shape model for a new species is not always possible due to the lack of 3D data. We present a method to capture new species using an articulated template and images of that species. In this work, we focus mainly on birds. Although birds represent almost twice the number of species as mammals, no accurate shape model is available. To capture a novel species, we first fit the articulated template to each training sample. By disentangling pose and shape, we learn a shape space that captures variation both among species and within each species from image evidence. We learn models of multiple species from the CUB dataset, and contribute new species-specific and multi-species shape models that are useful for downstream reconstruction tasks. Using a low-dimensional embedding, we show that our learned 3D shape space better reflects the phylogenetic relationships among birds than learned perceptual features.
38.Endless Loops: Detecting and Animating Periodic Patterns in Still Images ⬇️
We present an algorithm for producing a seamless animated loop from a single image. The algorithm detects periodic structures, such as the windows of a building or the steps of a staircase, and generates a non-trivial displacement vector field that maps each segment of the structure onto a neighboring segment along a user- or auto-selected main direction of motion. This displacement field is used, together with suitable temporal and spatial smoothing, to warp the image and produce the frames of a continuous animation loop. Our cinemagraphs are created in under a second on a mobile device. Over 140,000 users downloaded our app and exported over 350,000 cinemagraphs. Moreover, we conducted two user studies that show that users prefer our method for creating surreal and structured cinemagraphs compared to more manual approaches and compared to previous methods.
39.Efficient and Robust LiDAR-Based End-to-End Navigation ⬇️
Deep learning has been used to demonstrate end-to-end neural network learning for autonomous vehicle control from raw sensory input. While LiDAR sensors provide reliably accurate information, existing end-to-end driving solutions are mainly based on cameras since processing 3D data requires a large memory footprint and computation cost. On the other hand, increasing the robustness of these systems is also critical; however, even estimating the model's uncertainty is very challenging due to the cost of sampling-based methods. In this paper, we present an efficient and robust LiDAR-based end-to-end navigation framework. We first introduce Fast-LiDARNet that is based on sparse convolution kernel optimization and hardware-aware model design. We then propose Hybrid Evidential Fusion that directly estimates the uncertainty of the prediction from only a single forward pass and then fuses the control predictions intelligently. We evaluate our system on a full-scale vehicle and demonstrate lane-stable as well as navigation capabilities. In the presence of out-of-distribution events (e.g., sensor failures), our system significantly improves robustness and reduces the number of takeovers in the real world.
40.POCFormer: A Lightweight Transformer Architecture for Detection of COVID-19 Using Point of Care Ultrasound ⬇️
The rapid and seemingly endless expansion of COVID-19 can be traced back to the inefficiency and shortage of testing kits that offer accurate results in a timely manner. An emerging popular technique, which adopts improvements made in mobile ultrasound technology, allows for healthcare professionals to conduct rapid screenings on a large scale. We present an image-based solution that aims at automating the testing process which allows for rapid mass testing to be conducted with or without a trained medical professional that can be applied to rural environments and third world countries. Our contributions towards rapid large-scale testing include a novel deep learning architecture capable of analyzing ultrasound data that can run in real-time and significantly improve the current state-of-the-art detection accuracies using image-based COVID-19 detection.
41.Anchor-based Plain Net for Mobile Image Super-Resolution ⬇️
Along with the rapid development of real-world applications, higher requirements on the accuracy and efficiency of image super-resolution (SR) are brought forward. Though existing methods have achieved remarkable success, the majority of them demand plenty of computational resources and large amount of RAM, and thus they can not be well applied to mobile device. In this paper, we aim at designing efficient architecture for 8-bit quantization and deploy it on mobile device. First, we conduct an experiment about meta-node latency by decomposing lightweight SR architectures, which determines the portable operations we can utilize. Then, we dig deeper into what kind of architecture is beneficial to 8-bit quantization and propose anchor-based plain net (ABPN). Finally, we adopt quantization-aware training strategy to further boost the performance. Our model can outperform 8-bit quantized FSRCNN by nearly 2dB in terms of PSNR, while satisfying realistic needs at the same time. Code is avaliable at this https URL Jet/SR_Mobile_Quantization.
42.Covid-19 Detection from Chest X-ray and Patient Metadata using Graph Convolutional Neural Networks ⬇️
The novel corona virus (Covid-19) has introduced significant challenges due to its rapid spreading nature through respiratory transmission. As a result, there is a huge demand for Artificial Intelligence (AI) based quick disease diagnosis methods as an alternative to high demand tests such as Polymerase Chain Reaction (PCR). Chest X-ray (CXR) Image analysis is such cost-effective radiography technique due to resource availability and quick screening. But, a sufficient and systematic data collection that is required by complex deep leaning (DL) models is more difficult and hence there are recent efforts that utilize transfer learning to address this issue. Still these transfer learnt models suffer from lack of generalization and increased bias to the training dataset resulting poor performance for unseen data. Limited correlation of the transferred features from the pre-trained model to a specific medical imaging domain like X-ray and overfitting on fewer data can be reasons for this circumstance. In this work, we propose a novel Graph Convolution Neural Network (GCN) that is capable of identifying bio-markers of Covid-19 pneumonia from CXR images and meta information about patients. The proposed method exploits important relational knowledge between data instances and their features using graph representation and applies convolution to learn the graph data which is not possible with conventional convolution on Euclidean domain. The results of extensive experiments of proposed model on binary (Covid vs normal) and three class (Covid, normal, other pneumonia) classification problems outperform different benchmark transfer learnt models, hence overcoming the aforementioned drawbacks.
43.DPN-SENet:A self-attention mechanism neural network for detection and diagnosis of COVID-19 from chest x-ray images ⬇️
Background and Objective: The new type of coronavirus is also called COVID-19. It began to spread at the end of 2019 and has now spread across the world. Until October 2020, It has infected around 37 million people and claimed about 1 million lives. We propose a deep learning model that can help radiologists and clinicians use chest X-rays to diagnose COVID-19 cases and show the diagnostic features of pneumonia. Methods: The approach in this study is: 1) we propose a data enhancement method to increase the diversity of the data set, thereby improving the generalization performance of the model. 2) Our deep convolution neural network model DPN-SE adds a self-attention mechanism to the DPN network. The addition of a self-attention mechanism has greatly improved the performance of the network. 3) Use the Lime interpretable library to mark the feature regions on the X-ray medical image that helps doctors more quickly diagnose COVID-19 in people. Results: Under the same network model, the data with and without data enhancement is put into the model for training respectively. At last, comparing two experimental results: among the 10 network models with different structures, 7 network models have improved their effects after using data enhancement, with an average improvement of 1% in recognition accuracy. We propose that the accuracy and recall rates of the DPN-SE network are 93% and 98% of cases (COVID vs. pneumonia bacteria vs. viral pneumonia vs. normal). Compared with the original DPN, the respective accuracy is improved by 2%. Conclusion: The data augmentation method we used has achieved effective results on a small amount of data set, showing that a reasonable data augmentation method can improve the recognition accuracy without changing the sample size and model structure. Overall, the proposed method and model can effectively become a very useful tool for clinical radiologists.
44.Semantic segmentation of multispectral photoacoustic images using deep learning ⬇️
Photoacoustic imaging has the potential to revolutionise healthcare due to the valuable information on tissue physiology that is contained in multispectral photoacoustic measurements. Clinical translation of the technology requires conversion of the high-dimensional acquired data into clinically relevant and interpretable information. In this work, we present a deep learning-based approach to semantic segmentation of multispectral photoacoustic images to facilitate the interpretability of recorded images. Manually annotated multispectral photoacoustic imaging data are used as gold standard reference annotations and enable the training of a deep learning-based segmentation algorithm in a supervised manner. Based on a validation study with experimentally acquired data of healthy human volunteers, we show that automatic tissue segmentation can be used to create powerful analyses and visualisations of multispectral photoacoustic images. Due to the intuitive representation of high-dimensional information, such a processing algorithm could be a valuable means to facilitate the clinical translation of photoacoustic imaging.
45.FVC: A New Framework towards Deep Video Compression in Feature Space ⬇️
Learning based video compression attracts increasing attention in the past few years. The previous hybrid coding approaches rely on pixel space operations to reduce spatial and temporal redundancy, which may suffer from inaccurate motion estimation or less effective motion compensation. In this work, we propose a feature-space video coding network (FVC) by performing all major operations (i.e., motion estimation, motion compression, motion compensation and residual compression) in the feature space. Specifically, in the proposed deformable compensation module, we first apply motion estimation in the feature space to produce motion information (i.e., the offset maps), which will be compressed by using the auto-encoder style network. Then we perform motion compensation by using deformable convolution and generate the predicted feature. After that, we compress the residual feature between the feature from the current frame and the predicted feature from our deformable compensation module. For better frame reconstruction, the reference features from multiple previous reconstructed frames are also fused by using the non-local attention mechanism in the multi-frame feature fusion module. Comprehensive experimental results demonstrate that the proposed framework achieves the state-of-the-art performance on four benchmark datasets including HEVC, UVG, VTL and MCL-JCV.
46.Medical Image Segmentation using Squeeze-and-Expansion Transformers ⬇️
Medical image segmentation is important for computer-aided diagnosis. Good segmentation demands the model to see the big picture and fine details simultaneously, i.e., to learn image features that incorporate large context while keep high spatial resolutions. To approach this goal, the most widely used methods -- U-Net and variants, extract and fuse multi-scale features. However, the fused features still have small "effective receptive fields" with a focus on local image cues, limiting their performance. In this work, we propose Segtran, an alternative segmentation framework based on transformers, which have unlimited "effective receptive fields" even at high feature resolutions. The core of Segtran is a novel Squeeze-and-Expansion transformer: a squeezed attention block regularizes the self attention of transformers, and an expansion block learns diversified representations. Additionally, we propose a new positional encoding scheme for transformers, imposing a continuity inductive bias for images. Experiments were performed on 2D and 3D medical image segmentation tasks: optic disc/cup segmentation in fundus images (REFUGE'20 challenge), polyp segmentation in colonoscopy images, and brain tumor segmentation in MRI scans (BraTS'19 challenge). Compared with representative existing methods, Segtran consistently achieved the highest segmentation accuracy, and exhibited good cross-domain generalization capabilities.
47.Classifying concepts via visual properties ⬇️
We assume that substances in the world are represented by two types of concepts, namely substance concepts and classification concepts, the former instrumental to (visual) perception, the latter to (language based) classification. Based on this distinction, we introduce a general methodology for building lexico-semantic hierarchies of substance concepts, where nodes are annotated with the media, e.g.,videos or photos, from which substance concepts are extracted, and are associated with the corresponding classification concepts. The methodology is based on Ranganathan's original faceted approach, contextualized to the problem of classifying substance concepts. The key novelty is that the hierarchy is built exploiting the visual properties of substance concepts, while the linguistically defined properties of classification concepts are only used to describe substance concepts. The validity of the approach is exemplified by providing some highlights of an ongoing project whose goal is to build a large scale multimedia multilingual concept hierarchy.
48.Heterogeneous Contrastive Learning ⬇️
With the advent of big data across multiple high-impact applications, we are often facing the challenge of complex heterogeneity. The newly collected data usually consist of multiple modalities and characterized with multiple labels, thus exhibiting the co-existence of multiple types of heterogeneity. Although state-of-the-art techniques are good at modeling the complex heterogeneity with sufficient label information, such label information can be quite expensive to obtain in real applications, leading to sub-optimal performance using these techniques. Inspired by the capability of contrastive learning to utilize rich unlabeled data for improving performance, in this paper, we propose a unified heterogeneous learning framework, which combines both weighted unsupervised contrastive loss and weighted supervised contrastive loss to model multiple types of heterogeneity. We also provide theoretical analyses showing that the proposed weighted supervised contrastive loss is the lower bound of the mutual information of two samples from the same class and the weighted unsupervised contrastive loss is the lower bound of the mutual information between the hidden representation of two views of the same sample. Experimental results on real-world data sets demonstrate the effectiveness and the efficiency of the proposed method modeling multiple types of heterogeneity.
49.Robust partial Fourier reconstruction for diffusion-weighted imaging using a recurrent convolutional neural network ⬇️
Purpose: To develop an algorithm for robust partial Fourier (PF) reconstruction applicable to diffusion-weighted (DW) images with non-smooth phase variations.
Methods: Based on an unrolled proximal splitting algorithm, a neural network architecture is derived which alternates between data consistency operations and regularization implemented by recurrent convolutions. In order to exploit correlations, multiple repetitions of the same slice are jointly reconstructed under consideration of permutation-equivariance. The proposed method is trained on DW liver data of 60 volunteers and evaluated on retrospectively and prospectively sub-sampled data of different anatomies and resolutions. In addition, the benefits of using a recurrent network over other unrolling strategies is investigated.
Results: Conventional PF techniques can be significantly outperformed in terms of quantitative measures as well as perceptual image quality. The proposed method is able to generalize well to brain data with contrasts and resolution not present in the training set. The reduction in echo time (TE) associated with prospective PF-sampling enables DW imaging with higher signal. Also, the TE increase in acquisitions with higher resolution can be compensated for. It can be shown that unrolling by means of a recurrent network produced better results than using a weight-shared network or a cascade of networks.
Conclusion: This work demonstrates that robust PF reconstruction of DW data is feasible even at strong PF factors in applications with severe phase variations. Since the proposed method does not rely on smoothness priors of the phase but uses learned recurrent convolutions instead, artifacts of conventional PF methods can be avoided.
50.VOILA: Visual-Observation-Only Imitation Learning for Autonomous Navigation ⬇️
While imitation learning for vision based autonomous mobile robot navigation has recently received a great deal of attention in the research community, existing approaches typically require state action demonstrations that were gathered using the deployment platform. However, what if one cannot easily outfit their platform to record these demonstration signals or worse yet the demonstrator does not have access to the platform at all? Is imitation learning for vision based autonomous navigation even possible in such scenarios? In this work, we hypothesize that the answer is yes and that recent ideas from the Imitation from Observation (IfO) literature can be brought to bear such that a robot can learn to navigate using only ego centric video collected by a demonstrator, even in the presence of viewpoint mismatch. To this end, we introduce a new algorithm, Visual Observation only Imitation Learning for Autonomous navigation (VOILA), that can successfully learn navigation policies from a single video demonstration collected from a physically different agent. We evaluate VOILA in the photorealistic AirSim simulator and show that VOILA not only successfully imitates the expert, but that it also learns navigation policies that can generalize to novel environments. Further, we demonstrate the effectiveness of VOILA in a real world setting by showing that it allows a wheeled Jackal robot to successfully imitate a human walking in an environment using a video recorded using a mobile phone camera.
51.Exploring The Limits Of Data Augmentation For Retinal Vessel Segmentation ⬇️
Retinal Vessel Segmentation is important for diagnosis of various diseases. The research on retinal vessel segmentation focuses mainly on improvement of the segmentation model which is usually based on U-Net architecture. In our study we use the U-Net architecture and we rely on heavy data augmentation in order to achieve better performance. The success of the data augmentation relies on successfully addressing the problem of input images. By analyzing input images and performing the augmentation accordingly we show that the performance of the U-Net model can be increased dramatically. Results are reported using the most widely used retina dataset, DRIVE.
52.Generative Adversarial Neural Architecture Search ⬇️
Despite the empirical success of neural architecture search (NAS) in deep learning applications, the optimality, reproducibility and cost of NAS schemes remain hard to assess. In this paper, we propose Generative Adversarial NAS (GA-NAS) with theoretically provable convergence guarantees, promoting stability and reproducibility in neural architecture search. Inspired by importance sampling, GA-NAS iteratively fits a generator to previously discovered top architectures, thus increasingly focusing on important parts of a large search space. Furthermore, we propose an efficient adversarial learning approach, where the generator is trained by reinforcement learning based on rewards provided by a discriminator, thus being able to explore the search space without evaluating a large number of architectures. Extensive experiments show that GA-NAS beats the best published results under several cases on three public NAS benchmarks. In the meantime, GA-NAS can handle ad-hoc search constraints and search spaces. We show that GA-NAS can be used to improve already optimized baselines found by other NAS methods, including EfficientNet and ProxylessNAS, in terms of ImageNet accuracy or the number of parameters, in their original search space.