Search

Digital watermarking technology and watermark attack methods are mutually reinforcing and complementary. Currently, traditional watermark attack methods are relatively mature, but these traditional attack methods will inevitably damage the visual quality of original images (OIs). Therefore, this paper proposes a covert attack method called residual dense based imperceptible watermark attack network (RD-IWAN). First, this paper designs a watermark attack residual dense network (WARDN) based on the residual dense network (RDN), which can effectively remove the watermark information in the middle and low frequency features of the watermarked image (WMI). Second, to improve the attack ability of the network, this paper innovatively proposes a progressive preprocessing method based on the information enhancement preprocessing method. Concurrently, to ensure the imperceptibility of this watermark attack method, a comprehensive loss function that combines the perceptual loss and mean square error loss (MSE) of OI and attacked watermarked image (AWMI) is designed in this study. Finally, attack experiments are designed and performed on watermarks with different embedding strengths and sizes. Experimental results show that, compared to traditional attack methods, the watermark attack method proposed in this paper exhibits stronger attack ability and higher imperceptibility. [ABSTRACT FROM AUTHOR]

Since gait can be utilized to identify individuals from a far distance without their interaction and coordination, recently many gait recognition methods have been proposed. However, due to a real-world scenario of clothing changes, a degradation occurs for most of these methods. Thus in this paper, a more efficient gait recognition method is proposed to address the problem of clothing variances. First, part-based gait features are formulated from two different perspectives, i.e., the separated body parts that are more robust to clothing changes and the estimated human skeleton key-point regions. It is reasonable to formulate such features for cloth-changing gait recognition, because these two perspectives are both less vulnerable to clothing changes. Given that each feature has its own advantages and disadvantages, a more efficient gait feature is generated in this paper by assembling these two features together. Moreover, since local features are more discriminative than global features, in this paper more attention is focused on the local short-range features. Also, unlike most methods, in our method we treat the estimated key-point features as a set of word embeddings, and a transformer encoder is specifically used to learn the dependence of each correlative key-points. The robustness and effectiveness of our proposed method are certified by experiments on CASIA Gait Dataset B, and it has achieved the state-of-the-art performance on this dataset. [ABSTRACT FROM AUTHOR]

The content based image retrieval aims to find the similar images from a large scale dataset against a query image. Generally, the similarity between the representative features of the query image and dataset images is used to rank the images for retrieval. In early days, various hand designed feature descriptors have been investigated based on the visual cues such as color, texture, shape, etc. that represent the images. However, the deep learning has emerged as a dominating alternative of hand-designed feature engineering from a decade. It learns the features automatically from the data. This paper presents a comprehensive survey of deep learning based developments in the past decade for content based image retrieval. The categorization of existing state-of-the-art methods from different perspectives is also performed for greater understanding of the progress. The taxonomy used in this survey covers different supervision, different networks, different descriptor type and different retrieval type. A performance analysis is also performed using the state-of-the-art methods. The insights are also presented for the benefit of the researchers to observe the progress and to make the best choices. The survey presented in this paper will help in further research progress in image retrieval using deep learning. [ABSTRACT FROM AUTHOR]

Smoke detection in video is a challenging task because of the irregular shape of smoke, its complex motion state, which is affected by temperature, wind and other external factors, and background disturbances. Pixel-based foreground modeling method is a crucial step in many smoke detection systems and can be applied to efficiently focus on a certain object or a specific region to detect movements or anomalies. In video analysis, it is a natural idea to move the focus from the pixel-level foreground to the feature-level foreground. In this paper, the feature foreground is generated by the middle layer of a convolutional neural network (CNN) to guide the temporal modeling process for smoke objects. A novel temporal module called the Feature Foreground Module (FFM) is proposed to boost learning of a smoke temporal representation. Consider the problem of smoke analysis in video, we present a novel unifying approach, named an enhanced feature foreground network (EFFNet), that performs both smoke source prediction and detection. Efficient branch networks are designed in EFFNet, to predict the source mask and bounding boxes of smoke plumes in video. To the best of our knowledge, this is the first paper to study the source of smoke using deep learning methods. Finally, experiments on a realistic smoke dataset and a public dataset show that EFFNet method performs much better than do previous state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Video summarization (VS) is generally formulated as a subset selection problem where a set of representative keyframes or key segments is selected from an entire video frame set. Though many sparse subset selection based VS algorithms have been proposed in the past decade, most of them adopt linear sparse formulation in the explicit feature vector space of video frames, and don’t consider the local or global relationships among frames. In this paper, we first extend the conventional sparse subset selection for VS into kernel block sparse subset selection (KBS3) to utilize the advantage of kernel sparse coding and introduce a local inter-frame relationship through packing of frame blocks. Going a step further, we propose a similarity based block sparse subset selection (SB2S3) model by applying a specially designed transformation matrix on the KBS3 model in order to introduce a kind of global inter-frame relationship through the similarity. Finally, a greedy pursuit based algorithm is devised for the proposed NP-hard model optimization. The proposed SB2S3 has the following advantages: 1) through the similarity between each frame and any other frame, the global relationship among all frames can be considered; 2) through block sparse coding, the local relationship of adjacent frames is further considered; and 3) it has a wider application, since features can derive similarity, but not vice versa. It is believed that the effect of modeling such global and local relationships among frames in this paper, is similar to that of modeling the long-range and short-range dependencies among frames in deep learning based methods. Experimental results on three benchmark datasets have demonstrated that the proposed approach is superior to not only other sparse subset selection based VS methods but also most unsupervised deep-learning based VS methods. [ABSTRACT FROM AUTHOR]

Deep learning has greatly promoted the development of multi-view stereo in recent years. However, how to measure the reliability of the estimated depth map for practical applications and make reasonable depth hypothesis sampling for the cost volume building in the coarse-to-fine architecture are still unresolved crucial problems. To this end, an Uncertainty Guided multi-view Network (UGNet) is proposed in this paper. In order to enable the network to perceive the uncertainty, an uncertainty-aware loss function is introduced, which not only can infer uncertainty implicitly in an unsupervised manner but also can reduce the bad impact of high uncertainty regions and the erroneous labels in the training set during training. Moreover, an uncertainty-based depth hypothesis sampling strategy is further proposed to adaptively determine the depth search range of each pixel for finer stages, which helps to generate more rational depth intervals compared with other methods and build more compact cost volumes without redundancy. Experimental results on DTU dataset, BlendedMVS dataset, Tanks and Temples dataset and ETH3D high-res benchmark show that our method achieves promising reconstruction results compared with other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

This paper introduces deformable convolution in deep learning based joint denoising and demosaicing (JDD), which yields more adaptable representation and larger receptive fields in features extraction for a superior restoration performance. However, the deformable convolution generally leads to considerable computational load and irregular memory access bottleneck, limiting its extensive deployment on edge devices. To address this issue, we develop grouping strategy and assign independent offsets to each kernel group to reduce the computation latency while keeping the accuracy. Motivated by the exploration for aggregate distribution characteristics of deformable offsets, we present the offset sharing methodology to simplify the memory access complexity of deformable convolution. As for hardware acceleration, we specially design a novel deformable matrix multiplication workflow incorporated with a deformable memory mapping unit to boost the computational throughput. The verification experiments on FPGA demonstrate that the proposed deformable convolution based JDD can restore 4K Ultra High Definition (UHD) images at 70FPS and yields significant promotion in visual effect and objective quality assessment. [ABSTRACT FROM AUTHOR]

The visual storytelling (VIST) task aims at generating reasonable, human-like and coherent stories with the image streams as input. Although many deep learning models have achieved promising results, most of them do not directly leverage the sentiment information of stories. In this paper, we propose a sentiment-aware generative model for VIST called SentiStory. The key of SentiStory is a multi-layered sentiment extraction module (MLSEM). For a given image stream, the higher layer gives coarse-grained but accurate sentiments, while the lower layer of the MLSEM extracts fine-grained but usually unreliable ones. The two layers are combined strategically to generate coherent and rich visual sentiment concepts for the VIST task. Results from both automatic and human evaluations demonstrate that with the help of the MLSEM, SentiStory achieves improvement in generating more coherent and human-like stories. [ABSTRACT FROM AUTHOR]

This paper proposes a deep learning method for low-light image enhancement, which exploits the generation capability of Neural Networks (NNs) while requiring no training samples except the input image itself. Based on the Retinex decomposition model, the reflectance and illumination of a low-light image are parameterized by two untrained NNs. The ambiguity between the two layers is resolved by the discrepancy between the two NNs in terms of architecture and capacity, while the complex noise with spatially-varying characteristics is handled by an illumination-adaptive self-supervised denoising module. The enhancement is done by jointly optimizing the Retinex decomposition and the illumination adjustment. Extensive experiments show that the proposed method not only outperforms existing non-learning-based and unsupervised-learning-based methods, but also competes favorably with some supervised-learning-based methods in extreme low-light conditions. [ABSTRACT FROM AUTHOR]

Point clouds have attracted attention as a representation of an object’s surface. Deep generative models have typically used a continuous map from a dense set in a latent space to express their variations. However, a continuous map cannot adequately express the varying numbers of holes. That is, previous approaches disregarded the topological structure of point clouds. Furthermore, a point cloud comprises several subparts, making it difficult to express it using a continuous map. This paper proposes ChartPointFlow, a flow-based deep generative model that forms a map conditioned on a label. Similar to a manifold chart, a map conditioned on a label is assigned to a continuous subset of a point cloud. Thus, ChartPointFlow is able to maintain the topological structure with clear boundaries and holes, whereas previous approaches generated blurry point clouds with fuzzy holes. The experimental results show that ChartPointFlow achieves state-of-the-art performance in various tasks, including generation, reconstruction, upsampling, and segmentation. [ABSTRACT FROM AUTHOR]

The separation of the data capture and analysis in modern vision systems has led to a massive amount of data transfer between the end devices and cloud computers, resulting in long latency, slow response, and high power consumption. Efficient hardware architectures are under focused development to enable Artificial Intelligence (AI) at the resource-limited sensing devices. One of the most promising solutions is to enable Processing-in-Pixel (PIP) scheme. However, the conventional schemes suffer from the low fill-factor issue. This paper proposes a PIP based Complementary Metal-Oxide-Semiconductor (CMOS) sensor architecture, which allows convolution operation before the column readout circuit to significantly reduce the overall power consumption while improving the resource utilization of the succeeding deep learning accelerator. The simulation results show that the proposed architecture could support the computing efficiency up to 3.37 TOPS/W at the 8-bit weight configuration, which is four times as high as the conventional schemes after normalization. The transistors required for each pixel are only 3.5T, significantly improving the fill-factor. [ABSTRACT FROM AUTHOR]

Image captioning generates descriptions in a natural language for a given image. Due to its great potential for a wide range of applications, many deep learning based-methods have been proposed. The co-occurrence of words such as mouse and keyboard, constitutes commonsense knowledge, which is referred to as consensus. However, it is challenging to consider commonsense knowledge in producing captions that have rich, natural, and meaningful semantics. In this paper, a Vision-enhanced and Consensus-aware Transformer (VCT) is proposed to exploit both visual information and consensus knowledge for image captioning with three key components: a vision-enhanced encoder, consensus-aware knowledge representation generator, and consensus-aware decoder. The vision-enhanced encoder extends the vanilla self-attention module with a memory-based attention module and a visual perception module for learning better visual representation of an image. Specifically, the relationships between regions in an image and the image’s global context are leveraged with scene memory in the memory-based attention module. The visual perception module further enhances the correlation among neighboring tokens in both the spatial and channel-wise dimensions. To learn consensus-aware representations, a word correlation graph is constructed by computing the statistical co-occurrence between semantic concepts. Then consensus knowledge can be acquired using a graph convolutional network in the consensus-aware knowledge representation generator. Finally, such consensus knowledge is integrated into the consensus-aware decoder through consensus memory and a knowledge-based control module to produce a caption. Experimental results on two popular benchmark datasets (MSCOCO and Flickr30k) demonstrate that our proposed model achieves state-of-the-art performance. Extensive ablation studies also validate the effectiveness of each component. [ABSTRACT FROM AUTHOR]

Video grounding aims at localizing the temporal moment related to the given language description, which is very helpful to many cross-modal content understanding applications like visual question answering and sentence-video search. Existing approaches usually directly regress the temporal boundaries of an event described by a query sentence in the video sequence. This direct regression manner often encounters a large decision space due to diverse target events and variable video durations, leading to inaccurate localization as well as inefficient grounding. This paper presents an efficient framework termed from which to where to facilitate video grounding. The core idea is imitating the reading comprehension process to gradually narrow the decision space, in what we decompose the direct regression into two steps. The “which” step first roughly selects a candidate area by evaluating which video segment in the predefined set is closest to the ground truth. To this end, we formulate this step into a multi-choice reading comprehension problem and propose a criterion to select the best-matched segment. In this way, the excessive decision space is effectively reduced. The “where” step aims to precisely regress the temporal boundary of the selected video segment from the shrunk decision space. We thus introduce a triple-span representation for each candidate video segment to use the regional context for better boundary regression. The “which” and “where” steps can be combined into a unified framework and learned end-to-end, leading to an efficient video grounding system. Extensive experiments on Charades-STA, ActivityNet-Captions, and TACoS benchmarks clearly demonstrate the effectiveness of our framework. [ABSTRACT FROM AUTHOR]

In order to retain more feature information of local areas on a point cloud, local grouping and subsampling are the necessary data structuring steps in most hierarchical deep learning models. Due to the disorder nature of the points in a point cloud, the significant time cost may be consumed when grouping and subsampling the points, which consequently results in poor scalability. This paper proposes a fast data structuring method called PSNet (Point Structuring Net). PSNet transforms the spatial features of the points and matches them to the features of local areas in a point cloud. PSNet achieves grouping and sampling at the same time while the existing methods process sampling and grouping in two separate steps (such as using FPS plus kNN). PSNet performs feature transformation pointwise while the existing methods uses the spatial relationship among the points as the reference for grouping. Thanks to these features, PSNet has two important advantages: 1) the grouping and sampling results obtained by PSNet is stable and permutation invariant; and 2) PSNet can be easily parallelized. PSNet can replace the data structuring methods in the mainstream point cloud deep learning models in a plug-and-play manner. We have conducted extensive experiments. The results show that PSNet can improve the training and inference speed significantly while maintaining the model accuracy. [ABSTRACT FROM AUTHOR]

As a powerful statistical signal modeling technique, sparse representation has been widely used in various image restoration (IR) applications. The sparsity-based methods have achieved leading performance in the past few decades. However, in recent years it has been surpassed by other methods, especially the recent deep learning based methods. In this paper, we address the question that whether sparse representation can be competitive again. The way we answer this question is to redesign it with a deep architecture. To be specific, we propose an end-to-end deep architecture that follows the process of the sparse representation based IR. In particular, we learn a sparse convolutional dictionary to replace the traditional dictionary, and a convolutional neural network (CNN) denoising prior to replace the image prior. Through end-to-end training, the parameters in convolutional dictionary and CNN denoiser can be jointly optimized. Experimental results on several representative IR tasks, including image denoising, deblurring and super-resolution, demonstrate that the proposed deep network can achieve superior performance against state-of-the-art model-based and learning-based methods. [ABSTRACT FROM AUTHOR]

Models for image semantics understanding, such as deep learning (DL) models and mathematical models, are often trained on specific dataset or configured with specific parameters. When deploying such models on new tasks in a different test environment, it requires considerable effort to re-train the model or extensive expertise to tune the parameters. In this paper, we propose a smart reinforcement learning (RL) agent that could learn to tune parameters automatically to enhance model performance. The learning process is formulated as a generic control task for parameter adjustment, and applied to two use scenarios: (1) image attributes tuning to improve object detection performance on fixed DL model, and (2) parameter tuning of the mathematical model (Level Set) for image segmentation. We design a novel dynamic threshold mechanism in a multi-branch RL agent to effectively tune parameters of image qualities (for object detection) and Level Set models (for object segmentation). We conduct experiments on Pascal-VOC testing set, MS COCO validation set and a proprietary dataset of industrial components, where we achieve substantial improvement on object detection accuracy. We also perform experiments on the automatic parameter tuning of Level Set models. Results show that our method facilitates considerable performance improvement on public datasets compared with baseline method. [ABSTRACT FROM AUTHOR]

Modeling 3D visual saliency has received great attention due to the development of emerging 3D display technologies. Traditional methods relying on low-level features may not be efficient in interpreting 3D visual content from high-level semantic perspective. Despite numerous efforts dedicated to this area, existing 3D visual saliency detection methods do not necessarily excel in exploring the stereoscopic image saliency driven by the intra-view and inter-view dependencies among left and right views. In this paper, we propose a visual saliency detection method for stereoscopic images grounded on adaptive viewpoint feature enhancement via binocular vision. More specifically, the correlation among left and right views is investigated through a delicately designed binocular stereoscopic saliency feature aggregation module, enabling the generation of more representative saliency features towards binocular vision. Subsequently, to further aggregate the saliency features in multiple scales, we design a progressive attention-based saliency feature pyramid extraction module to effectively integrate the features from top-level to down-level based on the network hierarchy mechanism. The saliency maps are ultimately produced for stereoscopic images by evaluating the obtained saliency features. In addition, we create a stereoscopic image saliency dataset (SIS-3D) that includes 1086 stereoscopic image pairs with various content and their corresponding human eye fixation annotations, aiming to further facilitate the research on visual saliency detection for stereoscopic images. Extensive experiments demonstrate that our proposed method improves CC by an average of 4.02% compared to representative counterparts on the newly built saliency dataset and another publicly available dataset. [ABSTRACT FROM AUTHOR]

Instance-wise contrastive learning (Instance-CL), which learns to map similar instances closer and different instances farther apart in the embedding space, has achieved considerable progress in self-supervised video representation learning. However, canonical Instance-CL does not handle properly the temporal similarities between different videos, limiting the representation capabilities of learned models. This paper presents a novel two-stage framework that combines Instance-CL and unsupervised clustering to progressively learn desirable temporal representations with high intra-class compactness. Specifically, (a) we first introduce a new consistency-preserving sampling strategy to generate positive/negative pairs. Compared to the traditional sampling methods, our sampling strategy focuses more on motion dynamics, resulting in more temporal-related feature representations. (b) To further explore the temporal similarities between videos so as to encourage intra-class compactness, we set temporal representations extracted from Instance-CL as an initializer, and iteratively use k-means clustering to generate pseudo-labels for training the encoder. We term our method as Improved Instance-CL with Deep Clustering (ICDC) and apply it to two downstream tasks, including action recognition and video retrieval. Extensive experimental results show that ICDC gains considerable improvements compared to the existing self-supervised methods. [ABSTRACT FROM AUTHOR]

It has been recognized that texture patterns with abundant high-frequency components, such as grass and water, produce visual masking effects, and the distortion in textures is hard to be perceived by human eyes than structure regions. However, modern video codecs in a rate-distortion optimized manner usually consume a lot of bits to encode textures, leading to the insufficiency in perceptual coding performance. Nowadays, with the rapid development of deep learning, learning based texture synthesis methods have been proposed to replace the coding process of prediction residuals to reduce the rate cost. In this paper, we present a deep texture synthesizer named edge-based texture synthesis framework (ETSF). At encoder side, the framework detects texture regions by semantic and fidelity classification criteria, and the detected regions are quantized coarsely by the hybrid coding framework. In texture characterization, ETSF extracts low-level edge features representing pixel intensity variation. Feature processing tools are developed to remove the spatiotemporal redundancy of edges. The processed edge information is compressed and transmitted. To effectively recover textures, we design an edge-based texture synthesis generative adversarial network (ETSGAN) at the decoder of ETSF, which can incorporate edge information into convolutional layers and generate realistic textures. Experimental results on a collected texture dataset show that the proposed ETSF can achieve an average of -12.8%, -14.2% and -9.6% MOS BD-rate under lowdelay_B, lowdelay_P and random_access configurations of VVC coding, respectively. [ABSTRACT FROM AUTHOR]

Due to the rapid development of deep learning, the performance of salient object detection has been constantly refreshed. Nevertheless, it is still challenging for existing methods to distinguish the location of salient objects and retain fine structural details. In this paper, a novel progressive dual-attention residual network (PDRNet) is proposed to exploit two complementary attention maps to guide residual learning, thus progressively refining prediction in a coarse-to-fine manner. We design a dual-attention residual module (DRM) to achieve residual refinement with the help of the dual attention (DA) scheme. Specifically, an attention map and its corresponding reverse attention map are used to make the network be aware of learning residual details from the perspective of the salient and non-salient regions, thus utilizing their complementarity to correct the mistakes of object parts and boundary details. Besides, a hierarchical feature screening module (HFSM) is designed to capture more powerful global contextual knowledge for locating salient objects. It establishes cross-scale skip connections among multi-scale features and utilizes the intra-channel dependency of these scales to enhance information interaction and feature representation. Extensive experiments have proved that our proposed PDRNet performs favorably against 18 state-of-the-art competitors on five benchmark datasets, demonstrating the effectiveness and superiority of our method. [ABSTRACT FROM AUTHOR]

Deep autoencoder-based approaches have achieved significant improvements on restoring damaged images, yet they still suffer from artifacts due to the inadequate representation and inaccurate regularization of existing features. In this paper, we propose a dual-pyramidal inpainting framework called DPNet to address these two limitations, which seamlessly integrates sufficient feature learning and dynamic regularization within an autoencoder network. Specifically, to exhaustively extract multi-scale features, we adopt layer-wise pyramidal convolution in encoder, which provides an arbitrary combination pool of various receptive fields. Subsequently, to tackle the patch deterioration problem in previous cross-scale non-local schemes, we further propose a Pyramidal Attention Mechanism (PAM) in decoder to acquire finer patches directly from learned layers. Mutually benefited with pyramidal features extraction in encoder, the dissemination space for non-local pixels in our PAM is notably enlarged to pyramidal level, thus significantly benefiting the feature representation. Moreover, to avoid the mask error accumulation in existing works, a dynamic normalization mechanism utilizing the spatial mask information updated in encoder is introduced, which further ensures the feature integrity and consistency. Such a dual-pyramidal structure along with dynamic normalization significantly improve the inpainting quality, outperforming existing competitors. Comprehensive experiments conducted on three benchmark datasets demonstrate that our DPNet performs favorably against the state-of-the-arts. [ABSTRACT FROM AUTHOR]

We observed that remarkable and impressive performance on image-based human pose estimation have been achieved by deep Convolutional Neural Networks (CNN). Nevertheless, directly applying these image-based models on videos is not only computionally intensive, but also may cause jitter and loss. The main reason is that the image-based models purely focus on the local features of individual frames and totally ignore the temporal information among adjacent frames. Some existing methods are proposed to address the temporal coherency issue. However, these methods need to be designed carefully and cannot be combined with existing image-based methods. In this paper, we propose a simple yet effective module to refine the estimated pose by exploiting the temporal coherency among the heatmaps of adjacent frames, which can be easily inserted into image-based networks as a plug-in. We show that the temporal coherency issue among the heatmap frames could be re-formulated as a graph path selection optimization problem. Moreover, to speed up the refinement process, we propose a hierarchical graph optimization to achieve the refinement from coarse to fine. Experimental results on two large-scale video pose estimation benchmarks show that our module can improve the performance with little speed loss when combined with image-based methods as an efficient plug-in. [ABSTRACT FROM AUTHOR]

Multimodal, alias, guided, image restoration is the reconstruction of a degraded image from a target modality with the aid of a high quality image from another modality. A similar task is image fusion; it refers to merging images from different modalities into a composite image. Traditional approaches for multimodal image restoration and fusion include analytical methods that are computationally expensive at inference time. Recently developed deep learning methods have shown a great performance at a reduced computational cost; however, since these methods do not incorporate prior knowledge about the problem at hand, they result in a “black box” model, that is, one can hardly say what the model has learned. In this paper, we formulate multimodal image restoration and fusion as a coupled convolutional sparse coding problem, and adopt the Method of Multipliers (MM) for its solution. Then, we use the MM-based solution to design a convolutional neural network (CNN) encoder that follows the principle of deep unfolding. To address multimodal image restoration and fusion, we design two multimodal models which employ the proposed encoder followed by an appropriately designed decoder that maps the learned representations to the desired output. Unlike most existing deep learning designs comprising multiple encoding branches followed by a concatenation or a linear combination fusion block, the proposed design provides an efficient and structured way to fuse information at different stages of the network, providing representations that can lead to accurate image reconstruction. The proposed models are applied to three image restoration tasks, as well as two image fusion tasks. Quantitative and qualitative comparisons against various state-of-the-art analytical and deep learning methods corroborate the superior performance of the proposed framework. [ABSTRACT FROM AUTHOR]

With the widespread use of deep learning in single object tracking task, mainstream tracking algorithms treat tracking as a combined classification and regression problem. Classification aims at locating an arbitrary target, and regression aims at estimating the corresponding bounding box. In this paper, we focus on regression and propose a novel box estimation network, which consists of a transformer encoder target pyramid guide (TPG) and transformer decoder target pyramid spread (TPS). Specifically, the transformer encoder TPG is designed to generate fine-grained prior knowledge with explicit representation for template targets. In contrast to the raw transformer encoder, we capture the visual dependence through local-global self-attention and deem the multi-scale target regions as the “local” region. Using this fine-grained prior knowledge, we design the transformer decoder TPS to spread it to the subsequent search regions with high affinity to accurately estimate the bounding boxes. Considering that self-attention fails to model information interaction across channels between the template target and search regions, we develop a channel-wise cross-attention block within the TPS as compensation. Extensive experiments on the OTB100, UAV123, NFS, VOT2020, VOT2021, LaSOT, LaSOT_ext, TrackingNet and GOT-10k benchmarks show that the proposed box estimation network outperforms most existing box estimation methods. Furthermore, our trackers based on this estimation network exhibit a competitive performance against state-of-the-art trackers. [ABSTRACT FROM AUTHOR]

Discriminative correlation filter (DCF)-based methods have demonstrated superior performance in UAV tracking via fusing multiple types of features and updating models online. However, most DCF-based trackers simply cascade different features, failing to fully take advantage of their complementary strength. In addition, online update strategies are limited to using a single and fixed learning rate, which often leads to model degradation when suffering tracking challenges. In this paper, we present an Auto-Perceiving Correlation Filter (APCF) which explicitly models the target and context with a novel Target State and Background Perception (TSBP) feature. Concretely, we first propose a simple yet effective State Evaluation Metric (SEM) to estimate target states by analyzing the spatial distribution of responses. Based on SEM, we extract TSBP features by adaptively selecting effective features depending on the current target state. Accordingly, a new online model update strategy is also introduced to avoid model degradation. Moreover, we further introduce a perception regularization term to make the extracted feature emphasis more on the target rather than background. Extensive experiments on four widely-used UAV benchmarks have well demonstrated the superiority of the proposed method compared with both DCF and deep learning based trackers while running at a high speed of 76.7 FPS on a single CPU. In addition, APCF with deep features also performs favorably against state-of-the-art trackers. [ABSTRACT FROM AUTHOR]

Recognition of thyroid glands and tissues of the neck is vital for screening related diseases in ultrasound videos. This task is subjective, challenging, and dependent on the experience of sonographer in current clinical practice. The purpose is to develop a fully automated thyroid gland and tissues of neck recognition framework to assist doctors in distinguishing the boundaries of different tissues. In this paper, we propose a novel deep learning framework that consists of a feature extraction network, region proposal network, object detection head, and spatial pyramid RoIAlign-based segmentation head. Designed spatial pyramid RoIAlign can efficiently capture local and global context features, and aggregates the multiple context information that makes the result much more reliable. A large dataset is constructed to train the proposed method. The performance is evaluated using the COCO metrics. The experimental results demonstrate that the proposed deep learning method can effectively realize the automatic recognition of the thyroid gland and tissues of neck in ultrasound videos. Considering the clinical practical application scenarios, we developed an automatic recognition system of thyroid and neck tissue based on edge computing, which can expediently assist doctors in distinguishing the boundaries between different tissues. [ABSTRACT FROM AUTHOR]

The depth video contains large smooth contents with sharp edges. Since the deep learning-based color video orientated intra prediction methods pay no attention to the characteristics of depth video, they are unsuitable for optimizing the coding efficiency of depth video. In this paper, a multiple resolution prediction method with deep up-sampling is proposed to promote the coding efficiency of depth video. To efficiently encode the depth blocks of different complexity, the depth block is selectively encoded at different resolutions, including $\times 1$ , $\times 1$ /2, and $\times 1$ /4 resolutions. If the block is encoded with a low-resolution (LR), the resolution of reconstructed LR depth block is recovered by an up-sampling network. To constrain the quality of both reconstructed high-resolution depth block and its synthesized view, a view synthesis distortion guidance mechanism is proposed for the up-sampling network. In addition, a distillation-based lightweight up-sampling network is proposed to reduce the computational complexity. Experimental results demonstrate that the proposed multiple resolution prediction method obtains an average of 10.84% BD-rate saving in comparison with 3D-HEVC. [ABSTRACT FROM AUTHOR]

Image restoration includes various kinds of tasks, such as image denoising, image deraining and low-light image enhancement, etc. Due to the domain shift problem of current supervised methods, researchers tend to adopt unsupervised image restoration methods. However, fake color or blur image, insufficient restoration and missing semantic information are three common problems when utilizing these methods. In this paper, we propose a new hybrid loss named Quality-Task-Perception (QTP) to deal with these three problems simultaneously. Specifically, this hybrid loss includes three components: quality, task and perception. The quality part overcomes the fake color or blur image problem by enforcing image quality scores of the restored images and those of the unpaired clean images to be similar. For the task part, we tackle the insufficient restoration problem by proposing to apply a task probability network to convert the unsupervised image restoration into a supervised classification problem, and this task probability network is learned from our proposed pipeline. The perception part handles the missing semantic information by restricting the multi-scale phase consistency between the degraded image and its restored version. Comprehensive experiments on both supervised and unsupervised datasets in three image restoration tasks demonstrate the superiority of our proposed approach. [ABSTRACT FROM AUTHOR]

Low level image restoration is an integral component of modern artificial intelligence (AI) driven camera pipelines. Most of these frameworks are based on deep neural networks which present a massive computational overhead on resource constrained platform like a mobile phone. In this paper, we propose several lightweight low-level modules which can be used to create a computationally low cost variant of a given baseline model. Recent works for efficient neural networks design have mainly focused on classification. However, low-level image processing falls under the ‘image-to-image’ translation genre which requires some additional computational modules not present in classification. This paper seeks to bridge this gap by designing generic efficient modules which can replace essential components used in contemporary deep learning based image restoration networks. We also present and analyse our results highlighting the drawbacks of applying depthwise separable convolutional kernel (a popular method for efficient classification network) for sub-pixel convolution based upsampling (a popular upsampling strategy for low-level vision applications). This shows that concepts from domain of classification cannot always be seamlessly integrated into ‘image-to-image’ translation tasks. We extensively validate our findings on three popular tasks of image inpainting, denoising and super-resolution. Our results show that proposed networks consistently output visually similar reconstructions compared to full capacity baselines with significant reduction of parameters, memory footprint and execution speeds on contemporary mobile devices. [ABSTRACT FROM AUTHOR]

Detecting anomalies in videos is a fundamental issue in public security. The majority of existing deep learning methods often perform anomaly detection based on the behavior or the trajectory of a single target. However, due to the overlaps of the crowd and the low-resolution of monitoring images, the segmentation of population is hard to implement and the features cannot be learned thoroughly, which make the methods be easily disturbed by visual elements and thus may lead to false detection sometimes. To tackle these problems, we propose the influence-aware attention to learn the representative attributes of the whole crowd. Walking pedestrians can be divided into numbers of flows, and in this paper, we aim to measure the consistency of movement patterns in the same stream and the interactions between different streams. Meanwhile, great importance is given to the relation between pedestrians and the circumstance for certain anomalies occur as a result of environmental issues. Specifically, the influence-aware attention module is composed of the motion attention and the location attention, which is designed to quantify the relations in the scene from spatial and temporal aspects. For the lack of abnormal samples, we utilize a dual generator-based framework to learn interactions among normal scenes. Experimental results on six benchmarks verify the effectiveness and robustness of our proposed method. [ABSTRACT FROM AUTHOR]

Due to occlusion among views and temporal inconsistency in depth video, spatio-temporal distortion occurs in 3D synthesized video with depth image-based rendering. In this paper, we propose a deep Convolutional Neural Network (CNN)-based synthesized video denoising algorithm to reduce temporal flicker distortion and improve perceptual quality of 3D synthesized video. First, we analyze the spatio-temporal distortion, and model eliminating spatio-temporal distortion as a perceptual video denoising problem. Then, a deep learning-based synthesized video denoising network is proposed, in which a CNN-friendly spatio-temporal loss function is derived from a synthesized video quality metric and integrated with a single image denoising network architecture. Finally, specific schemes, i.e., specific Synthesized Video Denoising Networks (SynVD-Nets), and a general scheme, i.e., General SynVD-Net (GSynVD-Net), based on existing CNN-based denoising models, are developed to handle synthesized video with different distortion levels more effectively. Experimental results show that the proposed SynVD-Net and GSynVD-Net can outperform deep learning-based counterparts and conventional denoising methods, and significantly enhance perceptual quality of 3D synthesized video. [ABSTRACT FROM AUTHOR]

As one of the key technologies of Versatile Video Coding (VVC), a flexible quad-tree with a nested multi-type tree (QTMT) partition structure significantly improves the rate-distortion (RD) performance. However, this structure brings additional complexity due to the recursive search for the best partition type. Traditional fast partition methods in previous encoders, cannot adapt to this new complex structure, because it’s too complicated to predict each block size from one layer to another layer. Some indirect bottom-up designed methods are simple enough, but cannot predict specific split structures, making the acceleration capacity limited. Therefore, in this paper, we propose a learning-based approach to effectively predict the QTMT structure without having to heuristically explore the partitions of each layer. Firstly, we propose a hierarchy grid fully convolutional network (HG-FCN) framework, which concisely requires inference only once to obtain the entire partition information of the current CU and sub-CUs, and the inference is highly parallel. Secondly, we design a representation of complicated QTMT of CU partition in the form of hierarchy grid map (HGM), which can directly and effectively predict the specific hierarchical split structure. Lastly, a dual-threshold decision scheme is adopted to automatically control the trade-off between coding performance and complexity. Extensive experiments demonstrate the effectiveness of HG-FCN, which can reduce 51.15% $\sim ~65.53$ % complexity of VVC intra coding with negligible 1.17% $\sim ~2.19$ % BD-BR increase, superior to other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Deep learning networks have recently demonstrated yielded impressive progress for multi-exposure image fusion. However, how to restore realistic texture details while correcting color distortion is still a challenging problem to be solved. To alleviate the aforementioned issues, in this paper, we propose an attention-guided global-local adversarial learning network for fusing extreme exposure images in a coarse-to-fine manner. Firstly, the coarse fusion result is generated under the guidance of attention weight maps, which acquires the essential region of interest from both sides. Secondly, we formulate an edge loss function, along with a spatial feature transform layer, for refining the fusion process. So that it can take full use of the edge information to deal with blurry edges. Moreover, by incorporating global-local learning, our method can balance pixel intensity distribution and correct the color distortion on spatially varying source images from both image/patch perspectives. Such a global-local discriminator ensures all the local patches of the fused images align with realistic normal-exposure ones. Extensive experimental results on two publicly available datasets show that our method drastically outperforms state-of-the-art methods in visual inspection and objective analysis. Furthermore, sufficient ablation experiments prove that our method has significant advantages in generating high-quality fused results with appealing details, clear targets, and faithful color. Source code will be available at https://github.com/JinyuanLiu-CV/AGAL. [ABSTRACT FROM AUTHOR]

It is challenging to restore low-resolution (LR) images to super-resolution (SR) images with correct and clear details. Existing deep learning works almost neglect the inherent structural information of images, which acts as an important role for visual perception of SR results. In this paper, we design a hierarchical feature exploitation network to probe and preserve structural information in a multi-scale feature fusion manner. First, we propose a cross convolution upon traditional edge detectors to localize and represent edge features. Then, cross convolution blocks (CCBs) are designed with feature normalization and channel attention to consider the inherent correlations of features. Finally, we leverage multi-scale feature fusion group (MFFG) to embed the cross convolution blocks and develop the relations of structural features in different scales hierarchically, invoking a lightweight structure-preserving network named as Cross-SRN. Experimental results demonstrate the Cross-SRN achieves competitive or superior restoration performances against the state-of-the-art methods with accurate and clear structural details. Moreover, we set a criterion to select images with rich structural textures. The proposed Cross-SRN outperforms the state-of-the-art methods on the selected benchmark, which demonstrates that our network has a significant advantage in preserving edges. [ABSTRACT FROM AUTHOR]

Partial Label Learning (PLL) is a weakly supervised learning framework where each training instance is associated with more than one candidate label. This learning method is dedicated to finding out the true label for each training instance. Most of the current PLL algorithms directly disambiguate the candidate labels without correcting the disambiguated results, making the algorithms vulnerable to the influence of instances easily misjudged. In this paper, GraphDCN, an innovative disambiguation correction net with inductive graph representation learning model is proposed. GraphDCN consists of a disambiguation model and a correction model. For a given instance, the disambiguation model tries to fit its underlying ground-truth label through the candidate label distribution of the instances connected with the given one, while the correction model tries to maximize the distance between the disambiguated labels and non-candidate labels, and uses the label probability thresholds to correct the disambiguated labels that may be wrong. As the training goes on, both the disambiguation and correction models alternately and iteratively boost their performance. Moreover, when considering the implementation of the disambiguation model, a partial cross entropy formulation is proposed to estimate the ground-truth label loss by updating the ambiguity confidence matrix, which can be proved satisfying convergence in PLL. Simulation results reveal the overwhelming performance of GraphDCN. [ABSTRACT FROM AUTHOR]

Latent subspace learning aims to produce a latent representation for better reconstruction and classification from high-dimensional data through exploiting the optimal subspace. Current latent subspace learning methods commonly have three problems: 1) The discriminative property is ignored when learning the latent subspace, 2) The redundancy exists between the latent subspace and the prediction space, 3) There is no unified latent subspace that exploits knowledge jointly from the raw space, latent subspace, and label space. In this paper, we formulate the Joint Discriminative Latent Subspace Learning (JDLSL) problem to address these issues, and provide its optimization solution. JDLSL learns image representation from two aspects: a) the joint learning of latent subspaces for data reconstruction and prediction, b) the joint learning of label space and latent subspace for data reconstruction. To integrate knowledge from the joint learning, we organize the sparsity-induced latent subspace, where row-sparsity and column sparsity are simultaneously imposed. We provide the theoretical proof for the discriminativity learning ability of the sparsity-induced latent subspace. Extensive experiments and comparisons with the state-of-the-art showed that the proposed method has better performance. JDLSL shows a competitive performance with deep features compared to deep learning architectures, reflecting it potential integrating with deep learning. [ABSTRACT FROM AUTHOR]

Unlike natural images, the topology similarity among meshes can hardly be handled with classical deep learning because of their irregular structures. Parameterization provides a way to represent meshes in the form of geometry and normal images, which reflects the correlation between neighboring sample locations. Generative Adversarial Networks (GANs) can efficiently generate images without explicitly computing probability densities of the underlying distribution. However, existing GANs such as Coupled Generative Adversarial Network (CoGAN) generally have two drawbacks: (1) Inability to process unnatural images. (2) Insufficient exploration of the inherent relation between normal and the corresponding geometry image. To address these issues, this paper proposes an efficient method named Prediction-Compensation Generative Adversarial Network (PCGAN) to learn a joint distribution of both geometry and normal images, which aims for generating meshes with two GANs. The consistency of two GANs for the geometry and the normal is guaranteed by utilizing a sequence of prediction-compensation pairs. The sequence can estimate the normal image from the geometry image and compensate the geometry from normal progressively. Particularly, the prediction has a closed-form expression, which provides high estimation accuracy and reduces training complexity. Extensive experimental results on facial mesh generation indicate that our PCGAN outperforms CoGAN and other architectures in retaining the geometry of the faces and in generating realistic face meshes with rich facial attributes such as facial expression and morphology. Moreover, quantitative evaluations demonstrate our superior performance compared with the methods mentioned above. [ABSTRACT FROM AUTHOR]

Face aging aims to estimate aged facial textures given a certain face image. A number of 2D face-aging methods have been developed, but there have been few studies on 3D face aging, which would be valuable in several real-world applications. The lack of 3D face-aging data has had a significant impact on the development of 3D face aging, but we hypothesized that the large amounts of 2D face-aging data on the internet could be leveraged for 3D aged facial textures. In this paper, we propose a novel 3D aging framework, which we call UV-transformation texture estimation based on generative adversarial networks (UVTE-GAN), to achieve 3D face aging. Specifically, the proposed framework has three parts: 1) a 3D vertex and texture estimator, which accurately estimates the face’s spatial vertices and textures; 2) a texture-aging GAN, which is responsible for aging the estimated texture map via adversarial learning; and 3) a 2D & 3D rendering rebuilder, which recovers 2D & 3D faces using the estimated facial vertex map and aged facial texture map. In addition, we also design a plugin layer that allows us to train the whole model in an end-to-end manner. Experimental results demonstrate the effectiveness of the proposed method in synthesizing visually pleasing 3D aged face pictures, and state-of-the-art performance is achieved on several public datasets. [ABSTRACT FROM AUTHOR]

Background matting is a recently developed image matting approach, with applications to image and video editing. It refers to estimating both the alpha matte and foreground from a pair of images with and without foreground objects. Recent work has applied deep learning to background matting, with very promising performance achieved. However, existing deep models are supervised which require a large dataset with ground truth alpha mattes for training. To avoid the cost of data collection and possible bias in training data, this paper proposes a dataset-free unsupervised deep learning-based approach for background matting. Observing that the local smoothness of alpha matte can be well characterized by the untrained network prior called deep matte prior, we model the foreground and alpha matte using the priors encoded by two generative convolutional neural networks. To avoid possible overfitting during unsupervised learning, a two-stage learning scheme is developed which contains projection-based training and Bayesian post refinement. An alpha-matte-driven initialization scheme is also developed for performance boost. Even without calling external training data, the proposed approach provides competitive performance to recent supervised learning-based methods in the experiments. [ABSTRACT FROM AUTHOR]

Monocular depth estimation, as one of the fundamental tasks of computer vision, plays a crucial role in three-dimensional (3D) scene understanding and perception. Usually, deep learning methods recover monocular depth maps using continuous regression manners by minimizing the errors between the ground-truth depth and the predicted depth. However, fine depth features may not be fully captured through layer-by-layer coding, which is prone to low spatial resolution depth maps and insufficient details. Furthermore, it usually converges slowly and suffers from unsatisfactory results. To tackle these issues, we propose a novel model, named context-based ordinal regression network (CORNet), to reconstruct monocular depth maps in the ordinal regression manner with context information in this paper. Firstly, we put forward a novel context-based encoder with a feature transformation (FT) module to learn context information and details from inputs, and output multi-scale feature maps. Then, we design a boundary enhancement module (BEM) with a spatial attention mechanism following each operation of feature fusion, which captures boundary features in the scene to enhance the border depth. Finally, a feature optimization module (FOM) is designed to fuse and optimize the multi-scale features and boundary features to strengthen depth learning. What’s more, we introduce an ordinal weighted inference to predict depth maps from probabilities and discretization values. Experiments and results on two challenging datasets, KITTI and NYU Depth V2, demonstrate that our proposed CORNet can estimate monocular depth maps effectively and obtain superior performance in capturing geometric features over existing methods. [ABSTRACT FROM AUTHOR]

The great success of deep learning has boosted the fast development of video quality enhancement. However, existing methods mainly focus on enhancing the objective quality of compressed video, and ignore their perceptual quality that plays a key role in determining quality of experience (QoE) of videos. In this paper, we aim at enhancing the perceptual quality of compressed video. Our main observation is that perceptual quality enhancement mostly relies on recovering the high-frequency details with fine textures. Accordingly, we propose a novel generative adversarial network (GAN) based on multi-level wavelet packet transform (WPT), which is called multi-level wavelet-based GAN+ (MW-GAN+), to exploit high-frequency details for enhancing the perceptual quality of compressed video. In MW-GAN+, we first propose a multi-level wavelet pixel-adaptive (MWP) module to extract temporal information across video frames, such that frame similarity can be utilized in recovering high-frequency details. Then, a wavelet reconstruction network, consisting of wavelet-dense residual blocks (WDRB), is developed to recover high-frequency details in a multi-level manner for enhanced frame reconstruction. Finally, we develop a 3D discriminator to encourage temporal coherence with a 3D-CNN based architecture. Experimental results demonstrate the superiority of our method over state-of-the-art methods in enhancing the perceptual quality of compressed video. Our code is available at https://github.com/IceClear/MW-GAN. [ABSTRACT FROM AUTHOR]

Most of the existing blind omnidirectional image quality assessment (BOIQA) methods are based on data-driven approach where the end-to-end neural network or deep learning tools are mainly used for feature extraction. However, it usually lacks interpretability and is difficult to discover the perceptual mechanism behind. In this paper, from the perspective of perception modeling, we propose a novel multi-angle projection based BOIQA (MP-BOIQA) method. Considering the omnibearing and near eye display characteristics with head mounted display, multiple color cubemap projection images with respect to different viewpoints are grouped as the color omnidirectional distortion (COD) units so as to simulate the user’s viewing behavior in subjective quality assessment. In the designed multi-angle projection based feature extractor, tensor decomposition is implemented on each COD unit for dimensionality reduction, and piecewise exponential fitting is used to get the distribution of mean subtracted contrast normalized coefficients of the unit’s feature matrices in tensor domain. Finally, the extracted features are pooled with random forest. The experimental results on three omnidirectional image quality datasets show that the MP-BOIQA method can deliver highly competitive performance compared with some representative full-reference quality assessment methods, as well as some state-of-the-art BOIQA methods. [ABSTRACT FROM AUTHOR]

In the Few-Shot Learning (FSL), much of the related efforts only rely on the few available labeled samples (support set) building approach. However, the challenge is that the support set is easy-to-be-biased, so that they cannot be competent prototypes and are hard to represent the class distribution, leading to performance bottlenecks. In this paper, we propose to solve this obstacle by capturing the distribution of the unlabeled samples (query set). We propose two sampling methods: DeepSearch ($\cal DS$) and WideSearch ($\cal WS$). Both approaches are simple to implement and have no trainable parameters. They search the query samples near to the support set in different manners. Afterward, the statistic information is calculated, and we generate the latent samples according to it. The generated latent set is promising. First, it brings the query set distribution information to the classifier, which significantly improves the performance of the cross-entropy-based classifier. Second, it helps the support set become the better prototypes, which boosts the performance of the prototype-based classifier. Third, we find few latent samples are enough to boost the performance. Abundant experiments prove the proposed method achieves state-of-the-art performance on the few-shot tasks. Finally, rich ablation studies explain the compelling details of our approach. [ABSTRACT FROM AUTHOR]

Homography estimation is an important task in computer vision applications, such as image stitching, video stabilization, and camera calibration. Traditional homography estimation methods heavily depend on the quantity and distribution of feature correspondences, leading to poor robustness in low-texture scenes. The learning solutions, on the contrary, try to learn robust deep features but demonstrate unsatisfying performance in the scenes with low overlap rates. In this paper, we address these two problems simultaneously by designing a contextual correlation layer (CCL). The CCL can efficiently capture the long-range correlation within feature maps and can be flexibly used in a learning framework. In addition, considering that a single homography can not represent the complex spatial transformation in depth-varying images with parallax, we propose to predict multi-grid homography from global to local. Moreover, we equip our network with a depth perception capability, by introducing a novel depth-aware shape-preserved loss. Extensive experiments demonstrate the superiority of our method over state-of-the-art solutions in the synthetic benchmark dataset and real-world dataset. The codes and models will be available at https://github.com/nie-lang/Multi-Grid-Deep-Homography. [ABSTRACT FROM AUTHOR]

Edge-preserving image filtering is an essential task in computational photography and imaging. In this paper, we propose a simple yet effective global edge-preserving filter based on soft clustering, and we propose a novel soft clustering algorithm based on a restricted Gaussian mixture model. Given specified parameters, the soft clustering process is firstly performed on the image to derive the partition matrix, from which the affinity matrix is then constructed for filtering. The filtering output is calculated as the weighted average of the pixels in the local window, so the proposed filter could suppress the intensity shift artifacts that impede most global filters. Besides, the weights in the proposed filter are derived by clustering, which properly separates dissimilar pixels, so the proposed filter could handle the halo artifacts that haunt many local filters. Moreover, our filter provides flexible control over the amount of smoothing that is deficient in the deep learning-based filters. Besides the efficacy in smoothing, the proposed filter naturally has low computational complexity. Qualitative and quantitative results suggest that the proposed filter benefits various applications, including edge-preserving smoothing, image enhancing, flash/non-flash fusion, HDR tone mapping, and dehazing. [ABSTRACT FROM AUTHOR]

The core purpose of deep metric learning is to construct an embedding space, where objects belonging to the same class are gathered together and the ones from different classes are pushed apart. Most existing approaches typically insist to inter-class characteristics, e.g., class-level information or instance-level similarity, to obtain semantic relevance of data points and get a large margin between different classes in the embedding space. However, the intra-class characteristics, e.g., local manifold structure or relative relationship within the same class, are usually overlooked in the learning process. Hence the output embeddings have limitation in retrieving a good ranking result if existing multiple positive samples. And the local data structure of embedding space cannot be fully exploited since lack of relative ranking information. As a result, the model is prone to overfitting on a train set and get low generalization on the test set (unseen classes) when losing sight of intra-class variance. This paper presents a novel self-supervised synthesis ranking auxiliary framework, which captures intra-class characteristics as well as inter-class characteristics for better metric learning. Our method designs a synthetic samples generation of polar coordinates to generate measurable intra-class variance with different strength and diversity in the latent space, which can simulate the various local structure change of intra-class in the initial data domain. And then formulates a self-supervised learning procedure to fully exploit this property and preserve it in the embedding space. As a result, the learned embedding space not only keeps inter-class discrimination but also owns subtle intra-class diversity, leading to better global and local embedding structures. Extensive experiments on five benchmarks show that our method significantly improves and outperforms the state-of-the-art methods on the performances of both retrieval and ranking by 2%-4% (personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained from the IEEE by sending an email to pubs-permissions@ieee.org). [ABSTRACT FROM AUTHOR]

Over the past few years, deep convolutional neural network-based methods have made great progress in semantic segmentation of street scenes. Some recent methods align feature maps to alleviate the semantic gap between them and achieve high segmentation accuracy. However, they usually adopt the feature alignment modules with the same network configuration in the decoder and thus ignore the different roles of stages of the decoder during feature aggregation, leading to a complex decoder structure. Such a manner greatly affects the inference speed. In this paper, we present a novel Stage-aware Feature Alignment Network (SFANet) based on the encoder-decoder structure for real-time semantic segmentation of street scenes. Specifically, a Stage-aware Feature Alignment module (SFA) is proposed to align and aggregate two adjacent levels of feature maps effectively. In the SFA, by taking into account the unique role of each stage in the decoder, a novel stage-aware Feature Enhancement Block (FEB) is designed to enhance spatial details and contextual information of feature maps from the encoder. In this way, we are able to address the misalignment problem with a very simple and efficient multi-branch decoder structure. Moreover, an auxiliary training strategy is developed to explicitly alleviate the multi-scale object problem without bringing additional computational costs during the inference phase. Experimental results show that the proposed SFANet exhibits a good balance between accuracy and speed for real-time semantic segmentation of street scenes. In particular, based on ResNet-18, SFANet respectively obtains 78.1% and 74.7% mean of class-wise Intersection-over-Union (mIoU) at inference speeds of 37 FPS and 96 FPS on the challenging Cityscapes and CamVid test datasets by using only a single GTX 1080Ti GPU. [ABSTRACT FROM AUTHOR]

This paper considers deep visual recognition on long-tailed data. Under the long-tailed distribution, a small portion of the classes (head classes) occupy most training samples and the most classes (tail classes) only occupy relatively few samples. We observe that such long-tailed distribution significantly distorts the deeply-learned feature space, which consequentially compromises the deep visual recognition. Specifically, during training, each head class is prone to a relatively wide spatial distribution in the deep feature space, while each tail class is prone to a relatively small spatial distribution. In another word, the tail classes usually have much smaller spatial distribution than the head classes, distorting the overall feature space. In response, we propose to explicitly inflate the distribution of each tail class in the deep feature space, so that the tail classes will have comparable distribution range as the head classes. To this end, we replace each tail feature vector with a set of feature vectors on the fly. These feature vectors follow a probabilistic distribution learned from the head classes and yield a “feature cloud” surrounding the original tail feature. We show that the feature cloud effectively transfers the within-class diversity from the head classes onto the tail classes, maintaining an effect of probabilistic feature augmentation. An important advantage of the proposed feature cloud is that it is capable to bring general improvement to long-tailed visual recognition on two fundamental tasks, i.e., deep classification and deep representation learning, in spite of the significant differences between them. Extensive experiments on both deep metric learning benchmarks and deep image classification benchmarks validate the effectiveness of the proposed feature cloud. [ABSTRACT FROM AUTHOR]

In real-world applications, complete or incomplete multi-view data are common, which leads to the problem of generalized multi-view clustering. Recently, researchers attempt to learn the latent representation in the common subspace from heterogeneous data, which usually suffers from feature degeneration. Moreover, there are limited efforts on simultaneously revealing the underlying subspace structure and exploring the complementary information from incomplete multiple views. In this paper, we introduce a novel Generalized Multi-view Collaborative Subspace Clustering (GMCSC) framework to address the above issues, in which consensus subspace structure of all views and embedding subspaces for each view are jointly learned to benefit each other. Specifically, we develop a novel collaborative subspace learning strategy based on self-representation learning, which provides a brand-new way of pursuing the complete subspace structure directly from multi-view data. Furthermore, we explore complementary information by enforcing the consistency across different views and preserving the view-specific information of each view, which can alleviate the problem of feature degeneration and enhance the reasonability of using a consensus representation for multiple views. Experimental results on six benchmark datasets demonstrate that the proposed method can significantly outperform the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Decentralized learning has gained great popularity to improve learning efficiency and preserve data privacy. Each computing node makes equal contribution to collaboratively learn a Deep Learning model. The elimination of centralized Parameter Servers (PS) can effectively address many issues such as privacy, performance bottleneck and single-point-failure. However, how to achieve Byzantine Fault Tolerance in decentralized learning systems is rarely explored, although this problem has been extensively studied in centralized systems. In this paper, we present an in-depth study towards the Byzantine resilience of decentralized learning systems with two contributions. First, from the adversarial perspective, we theoretically illustrate that Byzantine attacks are more dangerous and feasible in decentralized learning systems: even one malicious participant can arbitrarily alter the models of other participants by sending carefully crafted updates to its neighbors. Second, from the defense perspective, we propose Ubar, a novel algorithm to enhance decentralized learning with Byzantine Fault Tolerance. Specifically, Ubar provides a Uniform Byzantine-resilient Aggregation Rule for benign nodes to select the useful parameter updates and filter out the malicious ones in each training iteration. It guarantees that each benign node in a decentralized system can train a correct model under very strong Byzantine attacks with an arbitrary number of faulty nodes. We conduct extensive experiments on standard image classification tasks and the results indicate that Ubar can effectively defeat both simple and sophisticated Byzantine attacks with higher performance efficiency than existing solutions. [ABSTRACT FROM AUTHOR]