1.Learning to See Through Obstructions with Layered Decomposition ⬇️
We present a learning-based approach for removing unwanted obstructions, such as window reflections, fence occlusions or raindrops, from a short sequence of images captured by a moving camera. Our method leverages the motion differences between the background and the obstructing elements to recover both layers. Specifically, we alternate between estimating dense optical flow fields of the two layers and reconstructing each layer from the flow-warped images via a deep convolutional neural network. The learning-based layer reconstruction allows us to accommodate potential errors in the flow estimation and brittle assumptions such as brightness consistency. We show that training on synthetically generated data transfers well to real images. Our results on numerous challenging scenarios of reflection and fence removal demonstrate the effectiveness of the proposed method.
2.Adversarial Generative Grammars for Human Activity Prediction ⬇️
In this paper we propose an adversarial generative grammar model for future prediction. The objective is to learn a model that explicitly captures temporal dependencies, providing a capability to forecast multiple, distinct future activities. Our adversarial grammar is designed so that it can learn stochastic production rules from the data distribution, jointly with its latent non-terminal representations. Being able to select multiple production rules during inference leads to different predicted outcomes, thus efficiently modeling many plausible futures. The adversarial generative grammar is evaluated on the Charades, MultiTHUMOS, Human3.6M, and 50 Salads datasets and on two activity prediction tasks: future 3D human pose prediction and future activity prediction. The proposed adversarial grammar outperforms the state-of-the-art approaches, being able to predict much more accurately and further in the future, than prior work.
3.Hardware-Centric AutoML for Mixed-Precision Quantization ⬇️
Model quantization is a widely used technique to compress and accelerate deep neural network (DNN) inference. Emergent DNN hardware accelerators begin to support mixed precision (1-8 bits) to further improve the computation efficiency, which raises a great challenge to find the optimal bitwidth for each layer: it requires domain experts to explore the vast design space trading off among accuracy, latency, energy, and model size, which is both time-consuming and sub-optimal. Conventional quantization algorithm ignores the different hardware architectures and quantizes all the layers in a uniform way. In this paper, we introduce the Hardware-Aware Automated Quantization (HAQ) framework which leverages the reinforcement learning to automatically determine the quantization policy, and we take the hardware accelerator's feedback in the design loop. Rather than relying on proxy signals such as FLOPs and model size, we employ a hardware simulator to generate direct feedback signals (latency and energy) to the RL agent. Compared with conventional methods, our framework is fully automated and can specialize the quantization policy for different neural network architectures and hardware architectures. Our framework effectively reduced the latency by 1.4-1.95x and the energy consumption by 1.9x with negligible loss of accuracy compared with the fixed bitwidth (8 bits) quantization. Our framework reveals that the optimal policies on different hardware architectures (i.e., edge and cloud architectures) under different resource constraints (i.e., latency, energy, and model size) are drastically different. We interpreted the implication of different quantization policies, which offer insights for both neural network architecture design and hardware architecture design.
4.A Boundary Based Out-of-Distribution Classifier for Generalized Zero-Shot Learning ⬇️
Generalized Zero-Shot Learning (GZSL) is a challenging topic that has promising prospects in many realistic scenarios. Using a gating mechanism that discriminates the unseen samples from the seen samples can decompose the GZSL problem to a conventional Zero-Shot Learning (ZSL) problem and a supervised classification problem. However, training the gate is usually challenging due to the lack of data in the unseen domain. To resolve this problem, in this paper, we propose a boundary based Out-of-Distribution (OOD) classifier which classifies the unseen and seen domains by only using seen samples for training. First, we learn a shared latent space on a unit hyper-sphere where the latent distributions of visual features and semantic attributes are aligned class-wisely. Then we find the boundary and the center of the manifold for each class. By leveraging the class centers and boundaries, the unseen samples can be separated from the seen samples. After that, we use two experts to classify the seen and unseen samples separately. We extensively validate our approach on five popular benchmark datasets including AWA1, AWA2, CUB, FLO and SUN. The experimental results show that our approach surpasses state-of-the-art approaches by a significant margin.
5.BREEDS: Benchmarks for Subpopulation Shift ⬇️
We develop a methodology for assessing the robustness of models to subpopulation shift---specifically, their ability to generalize to novel data subpopulations that were not observed during training. Our approach leverages the class structure underlying existing datasets to control the data subpopulations that comprise the training and test distributions. This enables us to synthesize realistic distribution shifts whose sources can be precisely controlled and characterized, within existing large-scale datasets. Applying this methodology to the ImageNet dataset, we create a suite of subpopulation shift benchmarks of varying granularity. We then validate that the corresponding shifts are tractable by obtaining human baselines for them. Finally, we utilize these benchmarks to measure the sensitivity of standard model architectures as well as the effectiveness of off-the-shelf train-time robustness interventions. Code and data available at this https URL .
6.KBGN: Knowledge-Bridge Graph Network for Adaptive Vision-Text Reasoning in Visual Dialogue ⬇️
Visual dialogue is a challenging task that needs to extract implicit information from both visual (image) and textual (dialogue history) contexts. Classical approaches pay more attention to the integration of the current question, vision knowledge and text knowledge, despising the heterogeneous semantic gaps between the cross-modal information. In the meantime, the concatenation operation has become de-facto standard to the cross-modal information fusion, which has a limited ability in information retrieval. In this paper, we propose a novel Knowledge-Bridge Graph Network (KBGN) model by using graph to bridge the cross-modal semantic relations between vision and text knowledge in fine granularity, as well as retrieving required knowledge via an adaptive information selection mode. Moreover, the reasoning clues for visual dialogue can be clearly drawn from intra-modal entities and inter-modal bridges. Experimental results on VisDial v1.0 and VisDial-Q datasets demonstrate that our model outperforms exiting models with state-of-the-art results.
7.GeLaTO: Generative Latent Textured Objects ⬇️
Accurate modeling of 3D objects exhibiting transparency, reflections and thin structures is an extremely challenging problem. Inspired by billboards and geometric proxies used in computer graphics, this paper proposes Generative Latent Textured Objects (GeLaTO), a compact representation that combines a set of coarse shape proxies defining low frequency geometry with learned neural textures, to encode both medium and fine scale geometry as well as view-dependent appearance. To generate the proxies' textures, we learn a joint latent space allowing category-level appearance and geometry interpolation. The proxies are independently rasterized with their corresponding neural texture and composited using a U-Net, which generates an output photorealistic image including an alpha map. We demonstrate the effectiveness of our approach by reconstructing complex objects from a sparse set of views. We show results on a dataset of real images of eyeglasses frames, which are particularly challenging to reconstruct using classical methods. We also demonstrate that these coarse proxies can be handcrafted when the underlying object geometry is easy to model, like eyeglasses, or generated using a neural network for more complex categories, such as cars.
8.TextRay: Contour-based Geometric Modeling for Arbitrary-shaped Scene Text Detection ⬇️
Arbitrary-shaped text detection is a challenging task due to the complex geometric layouts of texts such as large aspect ratios, various scales, random rotations and curve shapes. Most state-of-the-art methods solve this problem from bottom-up perspectives, seeking to model a text instance of complex geometric layouts with simple local units (e.g., local boxes or pixels) and generate detections with heuristic post-processings. In this work, we propose an arbitrary-shaped text detection method, namely TextRay, which conducts top-down contour-based geometric modeling and geometric parameter learning within a single-shot anchor-free framework. The geometric modeling is carried out under polar system with a bidirectional mapping scheme between shape space and parameter space, encoding complex geometric layouts into unified representations. For effective learning of the representations, we design a central-weighted training strategy and a content loss which builds propagation paths between geometric encodings and visual content. TextRay outputs simple polygon detections at one pass with only one NMS post-processing. Experiments on several benchmark datasets demonstrate the effectiveness of the proposed approach. The code is available at this https URL.
9.Exposing Deep-faked Videos by Anomalous Co-motion Pattern Detection ⬇️
Recent deep learning based video synthesis approaches, in particular with applications that can forge identities such as "DeepFake", have raised great security concerns. Therefore, corresponding deep forensic methods are proposed to tackle this problem. However, existing methods are either based on unexplainable deep networks which greatly degrades the principal interpretability factor to media forensic, or rely on fragile image statistics such as noise pattern, which in real-world scenarios can be easily deteriorated by data compression. In this paper, we propose an fully-interpretable video forensic method that is designed specifically to expose deep-faked videos. To enhance generalizability on videos with various content, we model the temporal motion of multiple specific spatial locations in the videos to extract a robust and reliable representation, called Co-Motion Pattern. Such kind of conjoint pattern is mined across local motion features which is independent of the video contents so that the instance-wise variation can also be largely alleviated. More importantly, our proposed co-motion pattern possesses both superior interpretability and sufficient robustness against data compression for deep-faked videos. We conduct extensive experiments to empirically demonstrate the superiority and effectiveness of our approach under both classification and anomaly detection evaluation settings against the state-of-the-art deep forensic methods.
10.TransNet V2: An effective deep network architecture for fast shot transition detection ⬇️
Although automatic shot transition detection approaches are already investigated for more than two decades, an effective universal human-level model was not proposed yet. Even for common shot transitions like hard cuts or simple gradual changes, the potential diversity of analyzed video contents may still lead to both false hits and false dismissals. Recently, deep learning-based approaches significantly improved the accuracy of shot transition detection using 3D convolutional architectures and artificially created training data. Nevertheless, one hundred percent accuracy is still an unreachable ideal. In this paper, we share the current version of our deep network TransNet V2 that reaches state-of-the-art performance on respected benchmarks. A trained instance of the model is provided so it can be instantly utilized by the community for a highly efficient analysis of large video archives. Furthermore, the network architecture, as well as our experience with the training process, are detailed, including simple code snippets for convenient usage of the proposed model and visualization of results.
11.Detecting Urban Dynamics Using Deep Siamese Convolutional Neural Networks ⬇️
Change detection is a fast-growing discipline in the areas of computer vision and remote sensing. In this work, we designed and developed a variant of convolutional neural network (CNN), known as Siamese CNN to extract features from pairs of Sentinel-2 temporal images of Mekelle city captured at different times and detect changes due to urbanization: buildings and roads. The detection capability of the proposed was measured in terms of overall accuracy (95.8), Kappa measure (72.5), recall (76.5), precision (77.7), F1 measure (77.1). The model has achieved a good performance in terms of most of these measures and can be used to detect changes in Mekelle and other cities at different time horizons undergoing urbanization.
12.Unified Representation Learning for Cross Model Compatibility ⬇️
We propose a unified representation learning framework to address the Cross Model Compatibility (CMC) problem in the context of visual search applications. Cross compatibility between different embedding models enables the visual search systems to correctly recognize and retrieve identities without re-encoding user images, which are usually not available due to privacy concerns. While there are existing approaches to address CMC in face identification, they fail to work in a more challenging setting where the distributions of embedding models shift drastically. The proposed solution improves CMC performance by introducing a light-weight Residual Bottleneck Transformation (RBT) module and a new training scheme to optimize the embedding spaces. Extensive experiments demonstrate that our proposed solution outperforms previous approaches by a large margin for various challenging visual search scenarios of face recognition and person re-identification.
13.Learning Stereo Matchability in Disparity Regression Networks ⬇️
Learning-based stereo matching has recently achieved promising results, yet still suffers difficulties in establishing reliable matches in weakly matchable regions that are textureless, non-Lambertian, or occluded. In this paper, we address this challenge by proposing a stereo matching network that considers pixel-wise matchability. Specifically, the network jointly regresses disparity and matchability maps from 3D probability volume through expectation and entropy operations. Next, a learned attenuation is applied as the robust loss function to alleviate the influence of weakly matchable pixels in the training. Finally, a matchability-aware disparity refinement is introduced to improve the depth inference in weakly matchable regions. The proposed deep stereo matchability (DSM) framework can improve the matching result or accelerate the computation while still guaranteeing the quality. Moreover, the DSM framework is portable to many recent stereo networks. Extensive experiments are conducted on Scene Flow and KITTI stereo datasets to demonstrate the effectiveness of the proposed framework over the state-of-the-art learning-based stereo methods.
14.DTVNet: Dynamic Time-lapse Video Generation via Single Still Image ⬇️
This paper presents a novel end-to-end dynamic time-lapse video generation framework, named DTVNet, to generate diversified time-lapse videos from a single landscape image, which are conditioned on normalized motion vectors. The proposed DTVNet consists of two submodules: \emph{Optical Flow Encoder} (OFE) and \emph{Dynamic Video Generator} (DVG). The OFE maps a sequence of optical flow maps to a \emph{normalized motion vector} that encodes the motion information inside the generated video. The DVG contains motion and content streams that learn from the motion vector and the single image respectively, as well as an encoder and a decoder to learn shared content features and construct video frames with corresponding motion respectively. Specifically, the \emph{motion stream} introduces multiple \emph{adaptive instance normalization} (AdaIN) layers to integrate multi-level motion information that are processed by linear layers. In the testing stage, videos with the same content but various motion information can be generated by different \emph{normalized motion vectors} based on only one input image. We further conduct experiments on Sky Time-lapse dataset, and the results demonstrate the superiority of our approach over the state-of-the-art methods for generating high-quality and dynamic videos, as well as the variety for generating videos with various motion information.
15.Transfer Learning for Protein Structure Classification and Function Inference at Low Resolution ⬇️
Structure determination is key to understanding protein function at a molecular level. Whilst significant advances have been made in predicting structure and function from amino acid sequence, researchers must still rely on expensive, time-consuming analytical methods to visualise detailed protein conformation. In this study, we demonstrate that it is possible to make accurate predictions of protein fold taxonomy from structures determined at low ($>$3 Angstroms) resolution, using a deep convolutional neural network trained on high-resolution structures ($\leq$3 Angstroms). Thus, we provide proof of concept for high-speed, low-cost protein structure classification at low resolution. We explore the relationship between the information content of the input image and the predictive power of the model, achieving state of the art performance on homologous superfamily prediction with maps of interatomic distance. Our findings contribute further evidence that inclusion of both amino acid alpha and beta carbon geometry in these maps improves classification performance over purely alpha carbon representations, and show that side-chain information may not be necessary for fine-grained structure predictions. Finally, we confirm that high-resolution, low-resolution and NMR-determined structures inhabit a common feature space, and thus provide a theoretical basis for mapping between domains to boost resolution.
16.HydraMix-Net: A Deep Multi-task Semi-supervised Learning Approach for Cell Detection and Classification ⬇️
Semi-supervised techniques have removed the barriers of large scale labelled set by exploiting unlabelled data to improve the performance of a model. In this paper, we propose a semi-supervised deep multi-task classification and localization approach HydraMix-Net in the field of medical imagining where labelling is time consuming and costly. Firstly, the pseudo labels are generated using the model's prediction on the augmented set of unlabelled image with averaging. The high entropy predictions are further sharpened to reduced the entropy and are then mixed with the labelled set for training. The model is trained in multi-task learning manner with noise tolerant joint loss for classification localization and achieves better performance when given limited data in contrast to a simple deep model. On DLBCL data it achieves 80% accuracy in contrast to simple CNN achieving 70% accuracy when given only 100 labelled examples.
17.Reinforced Wasserstein Training for Severity-Aware Semantic Segmentation in Autonomous Driving ⬇️
Semantic segmentation is important for many real-world systems, e.g., autonomous vehicles, which predict the class of each pixel. Recently, deep networks achieved significant progress w.r.t. the mean Intersection-over Union (mIoU) with the cross-entropy loss. However, the cross-entropy loss can essentially ignore the difference of severity for an autonomous car with different wrong prediction mistakes. For example, predicting the car to the road is much more servery than recognize it as the bus. Targeting for this difficulty, we develop a Wasserstein training framework to explore the inter-class correlation by defining its ground metric as misclassification severity. The ground metric of Wasserstein distance can be pre-defined following the experience on a specific task. From the optimization perspective, we further propose to set the ground metric as an increasing function of the pre-defined ground metric. Furthermore, an adaptively learning scheme of the ground matrix is proposed to utilize the high-fidelity CARLA simulator. Specifically, we follow a reinforcement alternative learning scheme. The experiments on both CamVid and Cityscapes datasets evidenced the effectiveness of our Wasserstein loss. The SegNet, ENet, FCN and Deeplab networks can be adapted following a plug-in manner. We achieve significant improvements on the predefined important classes, and much longer continuous playtime in our simulator.
18.Attention-based 3D Object Reconstruction from a Single Image ⬇️
Recently, learning-based approaches for 3D reconstruction from 2D images have gained popularity due to its modern applications, e.g., 3D printers, autonomous robots, self-driving cars, virtual reality, and augmented reality. The computer vision community has applied a great effort in developing functions to reconstruct the full 3D geometry of objects and scenes. However, to extract image features, they rely on convolutional neural networks, which are ineffective in capturing long-range dependencies. In this paper, we propose to substantially improve Occupancy Networks, a state-of-the-art method for 3D object reconstruction. For such we apply the concept of self-attention within the network's encoder in order to leverage complementary input features rather than those based on local regions, helping the encoder to extract global information. With our approach, we were capable of improving the original work in 5.05% of mesh IoU, 0.83% of Normal Consistency, and more than 10X the Chamfer-L1 distance. We also perform a qualitative study that shows that our approach was able to generate much more consistent meshes, confirming its increased generalization power over the current state-of-the-art.
19.Robust Long-Term Object Tracking via Improved Discriminative Model Prediction ⬇️
We propose an improved discriminative model prediction method for robust long-term tracking based on a pre-trained short-term tracker. The baseline pre-trained short-term tracker is SuperDiMP which combines the bounding-box regressor of PrDiMP with the standard DiMP classifier. Our tracker RLT-DiMP improves SuperDiMP in the following three aspects: (1) Uncertainty reduction using random erasing: To make our model robust, we exploit an agreement from multiple images after erasing random small rectangular areas as a certainty. And then, we correct the tracking state of our model accordingly. (2) Random search with spatio-temporal constraints: we propose a robust random search method with a score penalty applied to prevent the problem of sudden detection at a distance. (3) Background augmentation for more discriminative feature learning: We augment various backgrounds that are not included in the search area to train a more robust model in the background clutter. In experiments on the VOT-LT2020 benchmark dataset, the proposed method achieves comparable performance to the state-of-the-art long-term trackers. The source code is available at: this https URL.
20.A Study of Efficient Light Field Subsampling and Reconstruction Strategies ⬇️
Limited angular resolution is one of the main obstacles for practical applications of light fields. Although numerous approaches have been proposed to enhance angular resolution, view selection strategies have not been well explored in this area. In this paper, we study subsampling and reconstruction strategies for light fields. First, different subsampling strategies are studied with a fixed sampling ratio, such as row-wise sampling, column-wise sampling, or their combinations. Second, several strategies are explored to reconstruct intermediate views from four regularly sampled input views. The influence of the angular density of the input is also evaluated. We evaluate these strategies on both real-world and synthetic datasets, and optimal selection strategies are devised from our results. These can be applied in future light field research such as compression, angular super-resolution, and design of camera systems.
21.PROFIT: A Novel Training Method for sub-4-bit MobileNet Models ⬇️
4-bit and lower precision mobile models are required due to the ever-increasing demand for better energy efficiency in mobile devices. In this work, we report that the activation instability induced by weight quantization (AIWQ) is the key obstacle to sub-4-bit quantization of mobile networks. To alleviate the AIWQ problem, we propose a novel training method called PROgressive-Freezing Iterative Training (PROFIT), which attempts to freeze layers whose weights are affected by the instability problem stronger than the other layers. We also propose a differentiable and unified quantization method (DuQ) and a negative padding idea to support asymmetric activation functions such as h-swish. We evaluate the proposed methods by quantizing MobileNet-v1, v2, and v3 on ImageNet and report that 4-bit quantization offers comparable (within 1.48 % top-1 accuracy) accuracy to full precision baseline. In the ablation study of the 3-bit quantization of MobileNet-v3, our proposed method outperforms the state-of-the-art method by a large margin, 12.86 % of top-1 accuracy.
22.ClimAlign: Unsupervised statistical downscaling of climate variables via normalizing flows ⬇️
Downscaling is a landmark task in climate science and meteorology in which the goal is to use coarse scale, spatio-temporal data to infer values at finer scales. Statistical downscaling aims to approximate this task using statistical patterns gleaned from an existing dataset of downscaled values, often obtained from observations or physical models. In this work, we investigate the application of deep latent variable learning to the task of statistical downscaling. We present ClimAlign, a novel method for unsupervised, generative downscaling using adaptations of recent work in normalizing flows for variational inference. We evaluate the viability of our method using several different metrics on two datasets consisting of daily temperature and precipitation values gridded at low (1 degree latitude/longitude) and high (1/4 and 1/8 degree) resolutions. We show that our method achieves comparable predictive performance to existing supervised statistical downscaling methods while simultaneously allowing for both conditional and unconditional sampling from the joint distribution over high and low resolution spatial fields. We provide publicly accessible implementations of our method, as well as the baselines used for comparison, on GitHub.
23.Fast and Accurate Optical Flow based Depth Map Estimation from Light Fields ⬇️
Depth map estimation is a crucial task in computer vision, and new approaches have recently emerged taking advantage of light fields, as this new imaging modality captures much more information about the angular direction of light rays compared to common approaches based on stereoscopic images or multi-view. In this paper, we propose a novel depth estimation method from light fields based on existing optical flow estimation methods. The optical flow estimator is applied on a sequence of images taken along an angular dimension of the light field, which produces several disparity map estimates. Considering both accuracy and efficiency, we choose the feature flow method as our optical flow estimator. Thanks to its spatio-temporal edge-aware filtering properties, the different disparity map estimates that we obtain are very consistent, which allows a fast and simple aggregation step to create a single disparity map, which can then converted into a depth map. Since the disparity map estimates are consistent, we can also create a depth map from each disparity estimate, and then aggregate the different depth maps in the 3D space to create a single dense depth map.
24.Learning to Cluster under Domain Shift ⬇️
While unsupervised domain adaptation methods based on deep architectures have achieved remarkable success in many computer vision tasks, they rely on a strong assumption, i.e. labeled source data must be available. In this work we overcome this assumption and we address the problem of transferring knowledge from a source to a target domain when both source and target data have no annotations. Inspired by recent works on deep clustering, our approach leverages information from data gathered from multiple source domains to build a domain-agnostic clustering model which is then refined at inference time when target data become available. Specifically, at training time we propose to optimize a novel information-theoretic loss which, coupled with domain-alignment layers, ensures that our model learns to correctly discover semantic labels while discarding domain-specific features. Importantly, our architecture design ensures that at inference time the resulting source model can be effectively adapted to the target domain without having access to source data, thanks to feature alignment and self-supervision. We evaluate the proposed approach in a variety of settings, considering several domain adaptation benchmarks and we show that our method is able to automatically discover relevant semantic information even in presence of few target samples and yields state-of-the-art results on multiple domain adaptation benchmarks.
25.Real-Time Sign Language Detection using Human Pose Estimation ⬇️
We propose a lightweight real-time sign language detection model, as we identify the need for such a case in videoconferencing. We extract optical flow features based on human pose estimation and, using a linear classifier, show these features are meaningful with an accuracy of 80%, evaluated on the DGS Corpus. Using a recurrent model directly on the input, we see improvements of up to 91% accuracy, while still working under 4ms. We describe a demo application to sign language detection in the browser in order to demonstrate its usage possibility in videoconferencing applications.
26.R-MNet: A Perceptual Adversarial Network for Image Inpainting ⬇️
Facial image inpainting is a problem that is widely studied, and in recent years the introduction of Generative Adversarial Networks, has led to improvements in the field. Unfortunately some issues persists, in particular when blending the missing pixels with the visible ones. We address the problem by proposing a Wasserstein GAN combined with a new reverse mask operator, namely Reverse Masking Network (R-MNet), a perceptual adversarial network for image inpainting. The reverse mask operator transfers the reverse masked image to the end of the encoder-decoder network leaving only valid pixels to be inpainted. Additionally, we propose a new loss function computed in feature space to target only valid pixels combined with adversarial training. These then capture data distributions and generate images similar to those in the training data with achieved realism (realistic and coherent) on the output images. We evaluate our method on publicly available dataset, and compare with state-of-the-art methods. We show that our method is able to generalize to high-resolution inpainting task, and further show more realistic outputs that are plausible to the human visual system when compared with the state-of-the-art methods.
27.Fully-Automated Packaging Structure Recognition in Logistics Environments ⬇️
Within a logistics supply chain, a large variety of transported goods need to be handled, recognized and checked at many different network points. Often, huge manual effort is involved in recognizing or verifying packet identity or packaging structure, for instance to check the delivery for completeness. We propose a method for complete automation of packaging structure recognition: Based on a single image, one or multiple transport units are localized and, for each of these transport units, the characteristics, the total number and the arrangement of its packaging units is recognized. Our algorithm is based on deep learning models, more precisely convolutional neural networks for instance segmentation in images, as well as computer vision methods and heuristic components. We use a custom data set of realistic logistics images for training and evaluation of our method. We show that the solution is capable of correctly recognizing the packaging structure in approximately 85% of our test cases, and even more (91%) when focusing on most common package types.
28.Deep UAV Localization with Reference View Rendering ⬇️
This paper presents a framework for the localization of Unmanned Aerial Vehicles (UAVs) in unstructured environments with the help of deep learning. A real-time rendering engine is introduced that generates optical and depth images given a six Degrees-of-Freedom (DoF) camera pose, camera model, geo-referenced orthoimage, and elevation map. The rendering engine is embedded into a learning-based six-DoF Inverse Compositional Lucas-Kanade (ICLK) algorithm that is able to robustly align the rendered and real-world image taken by the UAV. To learn the alignment under environmental changes, the architecture is trained using maps spanning multiple years at high resolution. The evaluation shows that the deep 6DoF-ICLK algorithm outperforms its non-trainable counterparts by a large margin. To further support the research in this field, the real-time rendering engine and accompanying datasets are released along with this publication.
29.Sharp Multiple Instance Learning for DeepFake Video Detection ⬇️
With the rapid development of facial manipulation techniques, face forgery has received considerable attention in multimedia and computer vision community due to security concerns. Existing methods are mostly designed for single-frame detection trained with precise image-level labels or for video-level prediction by only modeling the inter-frame inconsistency, leaving potential high risks for DeepFake attackers. In this paper, we introduce a new problem of partial face attack in DeepFake video, where only video-level labels are provided but not all the faces in the fake videos are manipulated. We address this problem by multiple instance learning framework, treating faces and input video as instances and bag respectively. A sharp MIL (S-MIL) is proposed which builds direct mapping from instance embeddings to bag prediction, rather than from instance embeddings to instance prediction and then to bag prediction in traditional MIL. Theoretical analysis proves that the gradient vanishing in traditional MIL is relieved in S-MIL. To generate instances that can accurately incorporate the partially manipulated faces, spatial-temporal encoded instance is designed to fully model the intra-frame and inter-frame inconsistency, which further helps to promote the detection performance. We also construct a new dataset FFPMS for partially attacked DeepFake video detection, which can benefit the evaluation of different methods at both frame and video levels. Experiments on FFPMS and the widely used DFDC dataset verify that S-MIL is superior to other counterparts for partially attacked DeepFake video detection. In addition, S-MIL can also be adapted to traditional DeepFake image detection tasks and achieve state-of-the-art performance on single-frame datasets.
30.Rethinking Pseudo-LiDAR Representation ⬇️
The recently proposed pseudo-LiDAR based 3D detectors greatly improve the benchmark of monocular/stereo 3D detection task. However, the underlying mechanism remains obscure to the research community. In this paper, we perform an in-depth investigation and observe that the efficacy of pseudo-LiDAR representation comes from the coordinate transformation, instead of data representation itself. Based on this observation, we design an image based CNN detector named Patch-Net, which is more generalized and can be instantiated as pseudo-LiDAR based 3D detectors. Moreover, the pseudo-LiDAR data in our PatchNet is organized as the image representation, which means existing 2D CNN designs can be easily utilized for extracting deep features from input data and boosting 3D detection performance. We conduct extensive experiments on the challenging KITTI dataset, where the proposed PatchNet outperforms all existing pseudo-LiDAR based counterparts. Code has been made available at: this https URL.
31.Text as Neural Operator: Image Manipulation by Text Instruction ⬇️
In this paper, we study a new task that allows users to edit an input image using language instructions. In this image generation task, the inputs are a reference image and a text instruction that describes desired modifications to the input image. We propose a GAN-based method to tackle this problem. The key idea is to treat language as neural operators to locally modify the image feature. To this end, our model decomposes the generation process into finding where (spatial region) and how (text operators) to apply modifications. We show that the proposed model performs favorably against recent baselines on three datasets.
32.TCL: an ANN-to-SNN Conversion with Trainable Clipping Layers ⬇️
Spiking Neural Networks (SNNs) provide significantly lower power dissipationthan deep neural networks (DNNs), called as analog neural networks (ANNs) inthis work. Conventionally, SNNs have failed to arrive at the training accuraciesof ANNs. However, several recent researches have shown that this challenge canbe addressed by converting ANN to SNN instead of the direct training of SNNs.Nonetheless, the large latency of SNNs still limits their application, more prob-lematic for large size datasets such as Imagenet. It is challenging to overcome thisproblem since in SNNs, there is the trade-off relation between their accuracy and la-tency. In this work, we elegantly alleviate the problem by using a trainable clippinglayers, so called TCL. By combining the TCL with traditional data-normalizationtechniques, we respectively obtain 71.12% and 73.38% (on ImageNet) for VGG-16and RESNET-34 after the ANN to SNN conversion with the latency constraint of250 cycles.
33.Keypoint Autoencoders: Learning Interest Points of Semantics ⬇️
Understanding point clouds is of great importance. Many previous methods focus on detecting salient keypoints to identity structures of point clouds. However, existing methods neglect the semantics of points selected, leading to poor performance on downstream tasks. In this paper, we propose Keypoint Autoencoder, an unsupervised learning method for detecting keypoints. We encourage selecting sparse semantic keypoints by enforcing the reconstruction from keypoints to the original point cloud. To make sparse keypoint selection differentiable, Soft Keypoint Proposal is adopted by calculating weighted averages among input points. A downstream task of classifying shape with sparse keypoints is conducted to demonstrate the distinctiveness of our selected keypoints. Semantic Accuracy and Semantic Richness are proposed and our method gives competitive or even better performance than state of the arts on these two metrics.
34.Key-Nets: Optical Transformation Convolutional Networks for Privacy Preserving Vision Sensors ⬇️
Modern cameras are not designed with computer vision or machine learning as the target application. There is a need for a new class of vision sensors that are privacy preserving by design, that do not leak private information and collect only the information necessary for a target machine learning task. In this paper, we introduce key-nets, which are convolutional networks paired with a custom vision sensor which applies an optical/analog transform such that the key-net can perform exact encrypted inference on this transformed image, but the image is not interpretable by a human or any other key-net. We provide five sufficient conditions for an optical transformation suitable for a key-net, and show that generalized stochastic matrices (e.g. scale, bias and fractional pixel shuffling) satisfy these conditions. We motivate the key-net by showing that without it there is a utility/privacy tradeoff for a network fine-tuned directly on optically transformed images for face identification and object detection. Finally, we show that a key-net is equivalent to homomorphic encryption using a Hill cipher, with an upper bound on memory and runtime that scales quadratically with a user specified privacy parameter. Therefore, the key-net is the first practical, efficient and privacy preserving vision sensor based on optical homomorphic encryption.
35.Grasping Field: Learning Implicit Representations for Human Grasps ⬇️
In recent years, substantial progress has been made on robotic grasping of household objects. Yet, human grasps are still difficult to synthesize realistically. There are several key reasons: (1) the human hand has many degrees of freedom (more than robotic manipulators); (2) the synthesized hand should conform naturally to the object surface; and (3) it must interact with the object in a semantically and physical plausible manner. To make progress in this direction, we draw inspiration from the recent progress on learning-based implicit representations for 3D object reconstruction. Specifically, we propose an expressive representation for human grasp modelling that is efficient and easy to integrate with deep neural networks. Our insight is that every point in a three-dimensional space can be characterized by the signed distances to the surface of the hand and the object, respectively. Consequently, the hand, the object, and the contact area can be represented by implicit surfaces in a common space, in which the proximity between the hand and the object can be modelled explicitly. We name this 3D to 2D mapping as Grasping Field, parameterize it with a deep neural network, and learn it from data. We demonstrate that the proposed grasping field is an effective and expressive representation for human grasp generation. Specifically, our generative model is able to synthesize high-quality human grasps, given only on a 3D object point cloud. The extensive experiments demonstrate that our generative model compares favorably with a strong baseline. Furthermore, based on the grasping field representation, we propose a deep network for the challenging task of 3D hand and object reconstruction from a single RGB image. Our method improves the physical plausibility of the 3D hand-object reconstruction task over baselines.
36.Distributed Multi-agent Video Fast-forwarding ⬇️
In many intelligent systems, a network of agents collaboratively perceives the environment for better and more efficient situation awareness. As these agents often have limited resources, it could be greatly beneficial to identify the content overlapping among camera views from different agents and leverage it for reducing the processing, transmission and storage of redundant/unimportant video frames. This paper presents a consensus-based distributed multi-agent video fast-forwarding framework, named DMVF, that fast-forwards multi-view video streams collaboratively and adaptively. In our framework, each camera view is addressed by a reinforcement learning based fast-forwarding agent, which periodically chooses from multiple strategies to selectively process video frames and transmits the selected frames at adjustable paces. During every adaptation period, each agent communicates with a number of neighboring agents, evaluates the importance of the selected frames from itself and those from its neighbors, refines such evaluation together with other agents via a system-wide consensus algorithm, and uses such evaluation to decide their strategy for the next period. Compared with approaches in the literature on a real-world surveillance video dataset VideoWeb, our method significantly improves the coverage of important frames and also reduces the number of frames processed in the system.
37.Measures of Complexity for Large Scale Image Datasets ⬇️
Large scale image datasets are a growing trend in the field of machine learning. However, it is hard to quantitatively understand or specify how various datasets compare to each other - i.e., if one dataset is more complex or harder to ``learn'' with respect to a deep-learning based network. In this work, we build a series of relatively computationally simple methods to measure the complexity of a dataset. Furthermore, we present an approach to demonstrate visualizations of high dimensional data, in order to assist with visual comparison of datasets. We present our analysis using four datasets from the autonomous driving research community - Cityscapes, IDD, BDD and Vistas. Using entropy based metrics, we present a rank-order complexity of these datasets, which we compare with an established rank-order with respect to deep learning.
38.Locating Cephalometric X-Ray Landmarks with Foveated Pyramid Attention ⬇️
CNNs, initially inspired by human vision, differ in a key way: they sample uniformly, rather than with highest density in a focal point. For very large images, this makes training untenable, as the memory and computation required for activation maps scales quadratically with the side length of an image. We propose an image pyramid based approach that extracts narrow glimpses of the of the input image and iteratively refines them to accomplish regression tasks. To assist with high-accuracy regression, we introduce a novel intermediate representation we call 'spatialized features'. Our approach scales logarithmically with the side length, so it works with very large images. We apply our method to Cephalometric X-ray Landmark Detection and get state-of-the-art results.
39.Bipartite Graph Reasoning GANs for Person Image Generation ⬇️
We present a novel Bipartite Graph Reasoning GAN (BiGraphGAN) for the challenging person image generation task. The proposed graph generator mainly consists of two novel blocks that aim to model the pose-to-pose and pose-to-image relations, respectively. Specifically, the proposed Bipartite Graph Reasoning (BGR) block aims to reason the crossing long-range relations between the source pose and the target pose in a bipartite graph, which mitigates some challenges caused by pose deformation. Moreover, we propose a new Interaction-and-Aggregation (IA) block to effectively update and enhance the feature representation capability of both person's shape and appearance in an interactive way. Experiments on two challenging and public datasets, i.e., Market-1501 and DeepFashion, show the effectiveness of the proposed BiGraphGAN in terms of objective quantitative scores and subjective visual realness. The source code and trained models are available at this https URL.
40.Unsupervised Deep Metric Learning with Transformed Attention Consistency and Contrastive Clustering Loss ⬇️
Existing approaches for unsupervised metric learning focus on exploring self-supervision information within the input image itself. We observe that, when analyzing images, human eyes often compare images against each other instead of examining images individually. In addition, they often pay attention to certain keypoints, image regions, or objects which are discriminative between image classes but highly consistent within classes. Even if the image is being transformed, the attention pattern will be consistent. Motivated by this observation, we develop a new approach to unsupervised deep metric learning where the network is learned based on self-supervision information across images instead of within one single image. To characterize the consistent pattern of human attention during image comparisons, we introduce the idea of transformed attention consistency. It assumes that visually similar images, even undergoing different image transforms, should share the same consistent visual attention map. This consistency leads to a pairwise self-supervision loss, allowing us to learn a Siamese deep neural network to encode and compare images against their transformed or matched pairs. To further enhance the inter-class discriminative power of the feature generated by this network, we adapt the concept of triplet loss from supervised metric learning to our unsupervised case and introduce the contrastive clustering loss. Our extensive experimental results on benchmark datasets demonstrate that our proposed method outperforms current state-of-the-art methods for unsupervised metric learning by a large margin.
41.Visual Imitation Made Easy ⬇️
Visual imitation learning provides a framework for learning complex manipulation behaviors by leveraging human demonstrations. However, current interfaces for imitation such as kinesthetic teaching or teleoperation prohibitively restrict our ability to efficiently collect large-scale data in the wild. Obtaining such diverse demonstration data is paramount for the generalization of learned skills to novel scenarios. In this work, we present an alternate interface for imitation that simplifies the data collection process while allowing for easy transfer to robots. We use commercially available reacher-grabber assistive tools both as a data collection device and as the robot's end-effector. To extract action information from these visual demonstrations, we use off-the-shelf Structure from Motion (SfM) techniques in addition to training a finger detection network. We experimentally evaluate on two challenging tasks: non-prehensile pushing and prehensile stacking, with 1000 diverse demonstrations for each task. For both tasks, we use standard behavior cloning to learn executable policies from the previously collected offline demonstrations. To improve learning performance, we employ a variety of data augmentations and provide an extensive analysis of its effects. Finally, we demonstrate the utility of our interface by evaluating on real robotic scenarios with previously unseen objects and achieve a 87% success rate on pushing and a 62% success rate on stacking. Robot videos are available at this https URL.
42.A Review on Deep Learning Techniques for the Diagnosis of Novel Coronavirus (COVID-19) ⬇️
Novel coronavirus (COVID-19) outbreak, has raised a calamitous situation all over the world and has become one of the most acute and severe ailments in the past hundred years. The prevalence rate of COVID-19 is rapidly rising every day throughout the globe. Although no vaccines for this pandemic have been discovered yet, deep learning techniques proved themselves to be a powerful tool in the arsenal used by clinicians for the automatic diagnosis of COVID-19. This paper aims to overview the recently developed systems based on deep learning techniques using different medical imaging modalities like Computer Tomography (CT) and X-ray. This review specifically discusses the systems developed for COVID-19 diagnosis using deep learning techniques and provides insights on well-known data sets used to train these networks. It also highlights the data partitioning techniques and various performance measures developed by researchers in this field. A taxonomy is drawn to categorize the recent works for proper insight. Finally, we conclude by addressing the challenges associated with the use of deep learning methods for COVID-19 detection and probable future trends in this research area. This paper is intended to provide experts (medical or otherwise) and technicians with new insights into the ways deep learning techniques are used in this regard and how they potentially further works in combatting the outbreak of COVID-19.
43.3D FLAT: Feasible Learned Acquisition Trajectories for Accelerated MRI ⬇️
Magnetic Resonance Imaging (MRI) has long been considered to be among the gold standards of today's diagnostic imaging. The most significant drawback of MRI is long acquisition times, prohibiting its use in standard practice for some applications. Compressed sensing (CS) proposes to subsample the k-space (the Fourier domain dual to the physical space of spatial coordinates) leading to significantly accelerated acquisition. However, the benefit of compressed sensing has not been fully exploited; most of the sampling densities obtained through CS do not produce a trajectory that obeys the stringent constraints of the MRI machine imposed in practice. Inspired by recent success of deep learning based approaches for image reconstruction and ideas from computational imaging on learning-based design of imaging systems, we introduce 3D FLAT, a novel protocol for data-driven design of 3D non-Cartesian accelerated trajectories in MRI. Our proposal leverages the entire 3D k-space to simultaneously learn a physically feasible acquisition trajectory with a reconstruction method. Experimental results, performed as a proof-of-concept, suggest that 3D FLAT achieves higher image quality for a given readout time compared to standard trajectories such as radial, stack-of-stars, or 2D learned trajectories (trajectories that evolve only in the 2D plane while fully sampling along the third dimension). Furthermore, we demonstrate evidence supporting the significant benefit of performing MRI acquisitions using non-Cartesian 3D trajectories over 2D non-Cartesian trajectories acquired slice-wise.
44.Artificial Intelligence to Assist in Exclusion of Coronary Atherosclerosis during CCTA Evaluation of Chest-Pain in the Emergency Department: Preparing an Application for Real-World Use ⬇️
Coronary Computed Tomography Angiography (CCTA) evaluation of chest-pain patients in an Emergency Department (ED) is considered appropriate. While a negative CCTA interpretation supports direct patient discharge from an ED, labor-intensive analyses are required, with accuracy in jeopardy from distractions. We describe the development of an Artificial Intelligence (AI) algorithm and workflow for assisting interpreting physicians in CCTA screening for the absence of coronary atherosclerosis. The two-phase approach consisted of (1) Phase 1 - focused on the development and preliminary testing of an algorithm for vessel-centerline extraction classification in a balanced study population (n = 500 with 50% disease prevalence) derived by retrospective random case selection; and (2) Phase 2 - concerned with simulated-clinical Trialing of the developed algorithm on a per-case basis in a more real-world study population (n = 100 with 28% disease prevalence) from an ED chest-pain series. This allowed pre-deployment evaluation of the AI-based CCTA screening application which provides a vessel-by-vessel graphic display of algorithm inference results integrated into a clinically capable viewer. Algorithm performance evaluation used Area Under the Receiver-Operating-Characteristic Curve (AUC-ROC); confusion matrices reflected ground-truth vs AI determinations. The vessel-based algorithm demonstrated strong performance with AUC-ROC = 0.96. In both Phase 1 and Phase 2, independent of disease prevalence differences, negative predictive values at the case level were very high at 95%. The rate of completion of the algorithm workflow process (96% with inference results in 55-80 seconds) in Phase 2 depended on adequate image quality. There is potential for this AI application to assist in CCTA interpretation to help extricate atherosclerosis from chest-pain presentations.
45.AtrialJSQnet: A New Framework for Joint Segmentation and Quantification of Left Atrium and Scars Incorporating Spatial and Shape Information ⬇️
Left atrial (LA) and atrial scar segmentation from late gadolinium enhanced magnetic resonance imaging (LGE MRI) is an important task in clinical practice. %, to guide ablation therapy and predict treatment results for atrial fibrillation (AF) patients. The automatic segmentation is however still challenging, due to the poor image quality, the various LA shapes, the thin wall, and the surrounding enhanced regions. Previous methods normally solved the two tasks independently and ignored the intrinsic spatial relationship between LA and scars. In this work, we develop a new framework, namely AtrialJSQnet, where LA segmentation, scar projection onto the LA surface, and scar quantification are performed simultaneously in an end-to-end style. We propose a mechanism of shape attention (SA) via an explicit surface projection, to utilize the inherent correlation between LA and LA scars. In specific, the SA scheme is embedded into a multi-task architecture to perform joint LA segmentation and scar quantification. Besides, a spatial encoding (SE) loss is introduced to incorporate continuous spatial information of the target, in order to reduce noisy patches in the predicted segmentation. We evaluated the proposed framework on 60 LGE MRIs from the MICCAI2018 LA challenge. Extensive experiments on a public dataset demonstrated the effect of the proposed AtrialJSQnet, which achieved competitive performance over the state-of-the-art. The relatedness between LA segmentation and scar quantification was explicitly explored and has shown significant performance improvements for both tasks. The code and results will be released publicly once the manuscript is accepted for publication via this https URL.
46.The Umbrella software suite for automated asteroid detection ⬇️
We present the Umbrella software suite for asteroid detection, validation, identification and reporting. The current core of Umbrella is an open-source modular library, called Umbrella2, that includes algorithms and interfaces for all steps of the processing pipeline, including a novel detection algorithm for faint trails. Building on the library, we have also implemented a detection pipeline accessible both as a desktop program (ViaNearby) and via a web server (Webrella), which we have successfully used in near real-time data reduction of a few asteroid surveys on the Wide Field Camera of the Isaac Newton Telescope. In this paper we describe the library, focusing on the interfaces and algorithms available, and we present the results obtained with the desktop version on a set of well-curated fields used by the EURONEAR project as an asteroid detection benchmark.
47.Implanting Synthetic Lesions for Improving Liver Lesion Segmentation in CT Exams ⬇️
The success of supervised lesion segmentation algorithms using Computed Tomography (CT) exams depends significantly on the quantity and variability of samples available for training. While annotating such data constitutes a challenge itself, the variability of lesions in the dataset also depends on the prevalence of different types of lesions. This phenomenon adds an inherent bias to lesion segmentation algorithms that can be diminished, among different possibilities, using aggressive data augmentation methods. In this paper, we present a method for implanting realistic lesions in CT slices to provide a rich and controllable set of training samples and ultimately improving semantic segmentation network performances for delineating lesions in CT exams. Our results show that implanting synthetic lesions not only improves (up to around 12%) the segmentation performance considering different architectures but also that this improvement is consistent among different image synthesis networks. We conclude that increasing the variability of lesions synthetically in terms of size, density, shape, and position seems to improve the performance of segmentation models for liver lesion segmentation in CT slices.
48.Left Ventricular Wall Motion Estimation by Active Polynomials for Acute Myocardial Infarction Detection ⬇️
Echocardiogram (echo) is the earliest and the primary tool for identifying regional wall motion abnormalities (RWMA) in order to diagnose myocardial infarction (MI) or commonly known as heart attack. This paper proposes a novel approach, Active Polynomials, which can accurately and robustly estimate the global motion of the Left Ventricular (LV) wall from any echo in a robust and accurate way. The proposed algorithm quantifies the true wall motion occurring in LV wall segments so as to assist cardiologists diagnose early signs of an acute MI. It further enables medical experts to gain an enhanced visualization capability of echo images through color-coded segments along with their "maximum motion displacement" plots helping them to better assess wall motion and LV Ejection-Fraction (LVEF). The outputs of the method can further help echo-technicians to assess and improve the quality of the echocardiogram recording. A major contribution of this study is the first public echo database collection composed by physicians at the Hamad Medical Corporation Hospital in Qatar. The so-called HMC-QU database will serve as the benchmark for the forthcoming relevant studies. The results over the HMC-QU dataset show that the proposed approach can achieve high accuracy, sensitivity and precision in MI detection even though the echo quality is quite poor, and the temporal resolution is low.
49.Multi-modal segmentation of 3D brain scans using neural networks ⬇️
Purpose: To implement a brain segmentation pipeline based on convolutional neural networks, which rapidly segments 3D volumes into 27 anatomical structures. To provide an extensive, comparative study of segmentation performance on various contrasts of magnetic resonance imaging (MRI) and computed tomography (CT) scans. Methods: Deep convolutional neural networks are trained to segment 3D MRI (MPRAGE, DWI, FLAIR) and CT scans. A large database of in total 851 MRI/CT scans is used for neural network training. Training labels are obtained on the MPRAGE contrast and coregistered to the other imaging modalities. The segmentation quality is quantified using the Dice metric for a total of 27 anatomical structures. Dropout sampling is implemented to identify corrupted input scans or low-quality segmentations. Full segmentation of 3D volumes with more than 2 million voxels is obtained in less than 1s of processing time on a graphical processing unit. Results: The best average Dice score is found on
$T_1$ -weighted MPRAGE ($85.3\pm4.6,%$ ). However, for FLAIR ($80.0\pm7.1,%$ ), DWI ($78.2\pm7.9,%$ ) and CT ($79.1\pm 7.9,%$ ), good-quality segmentation is feasible for most anatomical structures. Corrupted input volumes or low-quality segmentations can be detected using dropout sampling. Conclusion: The flexibility and performance of deep convolutional neural networks enables the direct, real-time segmentation of FLAIR, DWI and CT scans without requiring$T_1$ -weighted scans.
50.Surgical Mask Detection with Convolutional Neural Networks and Data Augmentations on Spectrograms ⬇️
In many fields of research, labeled datasets are hard to acquire. This is where data augmentation promises to overcome the lack of training data in the context of neural network engineering and classification tasks. The idea here is to reduce model over-fitting to the feature distribution of a small under-descriptive training dataset. We try to evaluate such data augmentation techniques to gather insights in the performance boost they provide for several convolutional neural networks on mel-spectrogram representations of audio data. We show the impact of data augmentation on the binary classification task of surgical mask detection in samples of human voice (ComParE Challenge 2020). Also we consider four varying architectures to account for augmentation robustness. Results show that most of the baselines given by ComParE are outperformed.
51.PiNet: Attention Pooling for Graph Classification ⬇️
We propose PiNet, a generalised differentiable attention-based pooling mechanism for utilising graph convolution operations for graph level classification. We demonstrate high sample efficiency and superior performance over other graph neural networks in distinguishing isomorphic graph classes, as well as competitive results with state of the art methods on standard chemo-informatics datasets.
52.Extension of JPEG XS for Two-Layer Lossless Coding ⬇️
A two-layer lossless image coding method compatible with JPEG XS is proposed. JPEG XS is a new international standard for still image coding that has the characteristics of very low latency and very low complexity. However, it does not support lossless coding, although it can achieve visual lossless coding. The proposed method has a two-layer structure similar to JPEG XT, which consists of JPEG XS coding and a lossless coding method. As a result, it enables us to losslessly restore original images, while maintaining compatibility with JPEG XS.
53.Thick Cloud Removal of Remote Sensing Images Using Temporal Smoothness and Sparsity-Regularized Tensor Optimization ⬇️
In remote sensing images, the presence of thick cloud accompanying cloud shadow is a high probability event, which can affect the quality of subsequent processing and limit the scenarios of application. Hence, removing the thick cloud and cloud shadow as well as recovering the cloud-contaminated pixels is indispensable to make good use of remote sensing images. In this paper, a novel thick cloud removal method for remote sensing images based on temporal smoothness and sparsity-regularized tensor optimization (TSSTO) is proposed. The basic idea of TSSTO is that the thick cloud and cloud shadow are not only sparse but also smooth along the horizontal and vertical direction in images while the clean images are smooth along the temporal direction between images. Therefore, the sparsity norm is used to boost the sparsity of the cloud and cloud shadow, and unidirectional total variation (UTV) regularizers are applied to ensure the unidirectional smoothness. This paper utilizes alternation direction method of multipliers to solve the presented model and generate the cloud and cloud shadow element as well as the clean element. The cloud and cloud shadow element is purified to get the cloud area and cloud shadow area. Then, the clean area of the original cloud-contaminated images is replaced to the corresponding area of the clean element. Finally, the reference image is selected to reconstruct details of the cloud area and cloud shadow area using the information cloning method. A series of experiments are conducted both on simulated and real cloud-contaminated images from different sensors and with different resolutions, and the results demonstrate the potential of the proposed TSSTO method for removing cloud and cloud shadow from both qualitative and quantitative viewpoints.
54.SAFRON: Stitching Across the Frontier for Generating Colorectal Cancer Histology Images ⬇️
Synthetic images can be used for the development and evaluation of deep learning algorithms in the context of limited availability of annotations. In the field of computational pathology where histology images are large and visual context is crucial, synthesis of large tissue images via generative modeling is a challenging task due to memory and computing constraints hindering the generation of large images. To address this challenge, we propose a novel framework named as SAFRON to construct realistic large tissue image tiles from ground truth annotations while preserving morphological features and with minimal boundary artifacts at the seams. To this end, we train the proposed SAFRON framework based on conditional generative adversarial networks on large tissue image tiles from the Colorectal Adenocarcinoma Gland (CRAG) and DigestPath datasets. We demonstrate that our model can generate high quality and realistic image tiles of arbitrary large size after training it on relatively small image patches. We also show that training on synthetic data generated by SAFRON can significantly boost the performance of a standard algorithm for gland segmentation of colorectal cancer tissue images. Sample high resolution images generated using SAFRON are available at the URL:this https URL
55.Topic Adaptation and Prototype Encoding for Few-Shot Visual Storytelling ⬇️
Visual Storytelling~(VIST) is a task to tell a narrative story about a certain topic according to the given photo stream. The existing studies focus on designing complex models, which rely on a huge amount of human-annotated data. However, the annotation of VIST is extremely costly and many topics cannot be covered in the training dataset due to the long-tail topic distribution. In this paper, we focus on enhancing the generalization ability of the VIST model by considering the few-shot setting. Inspired by the way humans tell a story, we propose a topic adaptive storyteller to model the ability of inter-topic generalization. In practice, we apply the gradient-based meta-learning algorithm on multi-modal seq2seq models to endow the model the ability to adapt quickly from topic to topic. Besides, We further propose a prototype encoding structure to model the ability of intra-topic derivation. Specifically, we encode and restore the few training story text to serve as a reference to guide the generation at inference time. Experimental results show that topic adaptation and prototype encoding structure mutually bring benefit to the few-shot model on BLEU and METEOR metric. The further case study shows that the stories generated after few-shot adaptation are more relative and expressive.
56.ARPM-net: A novel CNN-based adversarial method with Markov Random Field enhancement for prostate and organs at risk segmentation in pelvic CT images ⬇️
Purpose: The research is to develop a novel CNN-based adversarial deep learning method to improve and expedite the multi-organ semantic segmentation of CT images, and to generate accurate contours on pelvic CT images. Methods: Planning CT and structure datasets for 110 patients with intact prostate cancer were retrospectively selected and divided for 10-fold cross-validation. The proposed adversarial multi-residual multi-scale pooling Markov Random Field (MRF) enhanced network (ARPM-net) implements an adversarial training scheme. A segmentation network and a discriminator network were trained jointly, and only the segmentation network was used for prediction. The segmentation network integrates a newly designed MRF block into a variation of multi-residual U-net. The discriminator takes the product of the original CT and the prediction/ground-truth as input and classifies the input into fake/real. The segmentation network and discriminator network can be trained jointly as a whole, or the discriminator can be used for fine-tuning after the segmentation network is coarsely trained. Multi-scale pooling layers were introduced to preserve spatial resolution during pooling using less memory compared to atrous convolution layers. An adaptive loss function was proposed to enhance the training on small or low contrast organs. The accuracy of modeled contours was measured with the Dice similarity coefficient (DSC), Average Hausdorff Distance (AHD), Average Surface Hausdorff Distance (ASHD), and relative Volume Difference (VD) using clinical contours as references to the ground-truth. The proposed ARPM-net method was compared to several stateof-the-art deep learning methods.
57.Spatio-temporal Attention Model for Tactile Texture Recognition ⬇️
Recently, tactile sensing has attracted great interest in robotics, especially for facilitating exploration of unstructured environments and effective manipulation. A detailed understanding of the surface textures via tactile sensing is essential for many of these tasks. Previous works on texture recognition using camera based tactile sensors have been limited to treating all regions in one tactile image or all samples in one tactile sequence equally, which includes much irrelevant or redundant information. In this paper, we propose a novel Spatio-Temporal Attention Model (STAM) for tactile texture recognition, which is the very first of its kind to our best knowledge. The proposed STAM pays attention to both spatial focus of each single tactile texture and the temporal correlation of a tactile sequence. In the experiments to discriminate 100 different fabric textures, the spatially and temporally selective attention has resulted in a significant improvement of the recognition accuracy, by up to 18.8%, compared to the non-attention based models. Specifically, after introducing noisy data that is collected before the contact happens, our proposed STAM can learn the salient features efficiently and the accuracy can increase by 15.23% on average compared with the CNN based baseline approach. The improved tactile texture perception can be applied to facilitate robot tasks like grasping and manipulation.
58.GANDALF: Generative Adversarial Networks with Discriminator-Adaptive Loss Fine-tuning for Alzheimer's Disease Diagnosis from MRI ⬇️
Positron Emission Tomography (PET) is now regarded as the gold standard for the diagnosis of Alzheimer's Disease (AD). However, PET imaging can be prohibitive in terms of cost and planning, and is also among the imaging techniques with the highest dosage of radiation. Magnetic Resonance Imaging (MRI), in contrast, is more widely available and provides more flexibility when setting the desired image resolution. Unfortunately, the diagnosis of AD using MRI is difficult due to the very subtle physiological differences between healthy and AD subjects visible on MRI. As a result, many attempts have been made to synthesize PET images from MR images using generative adversarial networks (GANs) in the interest of enabling the diagnosis of AD from MR. Existing work on PET synthesis from MRI has largely focused on Conditional GANs, where MR images are used to generate PET images and subsequently used for AD diagnosis. There is no end-to-end training goal. This paper proposes an alternative approach to the aforementioned, where AD diagnosis is incorporated in the GAN training objective to achieve the best AD classification performance. Different GAN lossesare fine-tuned based on the discriminator performance, and the overall training is stabilized. The proposed network architecture and training regime show state-of-the-art performance for three- and four- class AD classification tasks.
59.GANBERT: Generative Adversarial Networks with Bidirectional Encoder Representations from Transformers for MRI to PET synthesis ⬇️
Synthesizing medical images, such as PET, is a challenging task due to the fact that the intensity range is much wider and denser than those in photographs and digital renderings and are often heavily biased toward zero. Above all, intensity values in PET have absolute significance, and are used to compute parameters that are reproducible across the population. Yet, usually much manual adjustment has to be made in pre-/post- processing when synthesizing PET images, because its intensity ranges can vary a lot, e.g., between -100 to 1000 in floating point values. To overcome these challenges, we adopt the Bidirectional Encoder Representations from Transformers (BERT) algorithm that has had great success in natural language processing (NLP), where wide-range floating point intensity values are represented as integers ranging between 0 to 10000 that resemble a dictionary of natural language vocabularies. BERT is then trained to predict a proportion of masked values images, where its "next sentence prediction (NSP)" acts as GAN discriminator. Our proposed approach, is able to generate PET images from MRI images in wide intensity range, with no manual adjustments in pre-/post- processing. It is a method that can scale and ready to deploy.
60.Predicting Risk of Developing Diabetic Retinopathy using Deep Learning ⬇️
Diabetic retinopathy (DR) screening is instrumental in preventing blindness, but faces a scaling challenge as the number of diabetic patients rises. Risk stratification for the development of DR may help optimize screening intervals to reduce costs while improving vision-related outcomes. We created and validated two versions of a deep learning system (DLS) to predict the development of mild-or-worse ("Mild+") DR in diabetic patients undergoing DR screening. The two versions used either three-fields or a single field of color fundus photographs (CFPs) as input. The training set was derived from 575,431 eyes, of which 28,899 had known 2-year outcome, and the remaining were used to augment the training process via multi-task learning. Validation was performed on both an internal validation set (set A; 7,976 eyes; 3,678 with known outcome) and an external validation set (set B; 4,762 eyes; 2,345 with known outcome). For predicting 2-year development of DR, the 3-field DLS had an area under the receiver operating characteristic curve (AUC) of 0.79 (95%CI, 0.78-0.81) on validation set A. On validation set B (which contained only a single field), the 1-field DLS's AUC was 0.70 (95%CI, 0.67-0.74). The DLS was prognostic even after adjusting for available risk factors (p<0.001). When added to the risk factors, the 3-field DLS improved the AUC from 0.72 (95%CI, 0.68-0.76) to 0.81 (95%CI, 0.77-0.84) in validation set A, and the 1-field DLS improved the AUC from 0.62 (95%CI, 0.58-0.66) to 0.71 (95%CI, 0.68-0.75) in validation set B. The DLSs in this study identified prognostic information for DR development from CFPs. This information is independent of and more informative than the available risk factors.