Yang Wang

Zhongyan Zhang, Lei Wang, Yang Wang et al.

Quality feature representation is key to instance image retrieval. To attain it, existing methods usually resort to a deep model pre-trained on benchmark datasets or even fine-tune the model with a task-dependent labelled auxiliary dataset. Although achieving promising results, this approach is restricted by two issues: 1) the domain gap between benchmark datasets and the dataset of a given retrieval task; 2) the required auxiliary dataset cannot be readily obtained. In light of this situation, this work looks into a different approach which has not been well investigated for instance image retrieval previously: {can we learn feature representation \textit{specific to} a given retrieval task in order to achieve excellent retrieval?} Our finding is encouraging. By adding an object proposal generator to generate image regions for self-supervised learning, the investigated approach can successfully learn feature representation specific to a given dataset for retrieval. This representation can be made even more effective by boosting it with image similarity information mined from the dataset. As experimentally validated, such a simple ``self-supervised learning + self-boosting'' approach can well compete with the relevant state-of-the-art retrieval methods. Ablation study is conducted to show the appealing properties of this approach and its limitation on generalisation across datasets.

Yang Wang, Haipeng Liu, Zeqian Yi et al.

The state-of-the-art recommendation systems have shifted the attention to efficient recommendation, e.g., on-device recommendation, under memory constraints. To this end, the existing methods either focused on the lightweight embeddings for both users and items, or involved on-device systems enjoying the compact embeddings to enhance reusability and reduces space complexity. However, they focus solely on the coarse granularity of embedding, while overlook the fine-grained semantic nuances, to adversarially downgrade the efficacy of meta-embeddings in capturing the intricate relationship over both user and item, consequently resulting into the suboptimal recommendations. In this paper, we aim to study how the meta-embedding can efficiently learn varied grained semantics, together with how the fine-grained meta-embedding can strengthen the representation of coarse-grained meta-embedding. To answer these questions, we develop a novel graph neural networks (GNNs) based recommender where each user and item serves as the node, linked directly to coarse-grained virtual nodes and indirectly to fine-grained virtual nodes, ensuring different grained semantic learning, while disclosing: 1) In contrast to coarse-grained semantics, fine-grained semantics are well captured through sparse meta-embeddings, which adaptively 2) balance the embedding uniqueness and memory constraint. Additionally, the initialization method come up upon SparsePCA, along with a soft thresholding activation function to render the sparseness of the meta-embeddings. We propose a weight bridging update strategy that focuses on matching each coarse-grained meta-embedding with several fine-grained meta-embeddings based on the users/items' semantics. Extensive experiments substantiate our method's superiority over existing baselines. Our code is available at https://github.com/htyjers/C2F-MetaEmbed.

Yuqian Zhou, Jianbo Jiao, Haibin Huang et al.

Discriminative learning-based image denoisers have achieved promising performance on synthetic noises such as Additive White Gaussian Noise (AWGN). The synthetic noises adopted in most previous work are pixel-independent, but real noises are mostly spatially/channel-correlated and spatially/channel-variant. This domain gap yields unsatisfied performance on images with real noises if the model is only trained with AWGN. In this paper, we propose a novel approach to boost the performance of a real image denoiser which is trained only with synthetic pixel-independent noise data dominated by AWGN. First, we train a deep model that consists of a noise estimator and a denoiser with mixed AWGN and Random Value Impulse Noise (RVIN). We then investigate Pixel-shuffle Down-sampling (PD) strategy to adapt the trained model to real noises. Extensive experiments demonstrate the effectiveness and generalization of the proposed approach. Notably, our method achieves state-of-the-art performance on real sRGB images in the DND benchmark among models trained with synthetic noises. Codes are available at https://github.com/yzhouas/PD-Denoising-pytorch.

Daoyu Lin, Guangluan Xu, Xiaoke Wang et al.

Cloud-based overlays are often present in optical remote sensing images, thus limiting the application of acquired data. Removing clouds is an indispensable pre-processing step in remote sensing image analysis. Deep learning has achieved great success in the field of remote sensing in recent years, including scene classification and change detection. However, deep learning is rarely applied in remote sensing image removal clouds. The reason is the lack of data sets for training neural networks. In order to solve this problem, this paper first proposed the Remote sensing Image Cloud rEmoving dataset (RICE). The proposed dataset consists of two parts: RICE1 contains 500 pairs of images, each pair has images with cloud and cloudless size of 512*512; RICE2 contains 450 sets of images, each set contains three 512*512 size images. , respectively, the reference picture without clouds, the picture of the cloud and the mask of its cloud. The dataset is freely available at \url{https://github.com/BUPTLdy/RICE_DATASET}.

Chenxu Luo, Zhenheng Yang, Peng Wang et al.

Learning to estimate 3D geometry in a single frame and optical flow from consecutive frames by watching unlabeled videos via deep convolutional network has made significant progress recently. Current state-of-the-art (SoTA) methods treat the two tasks independently. One typical assumption of the existing depth estimation methods is that the scenes contain no independent moving objects. while object moving could be easily modeled using optical flow. In this paper, we propose to address the two tasks as a whole, i.e. to jointly understand per-pixel 3D geometry and motion. This eliminates the need of static scene assumption and enforces the inherent geometrical consistency during the learning process, yielding significantly improved results for both tasks. We call our method as "Every Pixel Counts++" or "EPC++". Specifically, during training, given two consecutive frames from a video, we adopt three parallel networks to predict the camera motion (MotionNet), dense depth map (DepthNet), and per-pixel optical flow between two frames (OptFlowNet) respectively. The three types of information are fed into a holistic 3D motion parser (HMP), and per-pixel 3D motion of both rigid background and moving objects are disentangled and recovered. Comprehensive experiments were conducted on datasets with different scenes, including driving scenario (KITTI 2012 and KITTI 2015 datasets), mixed outdoor/indoor scenes (Make3D) and synthetic animation (MPI Sintel dataset). Performance on the five tasks of depth estimation, optical flow estimation, odometry, moving object segmentation and scene flow estimation shows that our approach outperforms other SoTA methods. Code will be available at: https://github.com/chenxuluo/EPC.

Lei Sun, Andrea Alfarano, Peiqi Duan et al.

This paper presents an overview of NTIRE 2025 the First Challenge on Event-Based Image Deblurring, detailing the proposed methodologies and corresponding results. The primary goal of the challenge is to design an event-based method that achieves high-quality image deblurring, with performance quantitatively assessed using Peak Signal-to-Noise Ratio (PSNR). Notably, there are no restrictions on computational complexity or model size. The task focuses on leveraging both events and images as inputs for single-image deblurring. A total of 199 participants registered, among whom 15 teams successfully submitted valid results, offering valuable insights into the current state of event-based image deblurring. We anticipate that this challenge will drive further advancements in event-based vision research.

Xin Xu, Shizhe Diao, Can Yang et al.

Chain-of-Thought (CoT) prompting has marked a significant advancement in enhancing the reasoning capabilities of large language models (LLMs). Previous studies have developed various extensions of CoT, which focus primarily on enhancing end-task performance. In addition, there has been research on assessing the quality of reasoning chains in CoT. This raises an intriguing question: Is it possible to predict the accuracy of LLM outputs by scrutinizing the reasoning chains they generate? To answer this research question, we introduce a benchmark, R2PE, designed specifically to explore the relationship between reasoning chains and performance in various reasoning tasks spanning five different domains. This benchmark aims to measure the falsehood of the final output of LLMs based on the reasoning steps. To make full use of information in multiple reasoning chains, we propose the process discernibility score (PDS) framework that beats the answer-checking baseline by a large margin. Concretely, this resulted in an average of $5.1\%$ increase in the F1 score and $2.97\%$ improvement in AUC-PR across all 45 subsets within R2PE. We further demonstrate our PDS's efficacy in advancing open-domain QA accuracy.

Lili Zhao, Qi Liu, Wei Chen et al.

Empirical Risk Minimization (ERM) models often rely on spurious correlations between features and labels during the learning process, leading to shortcut learning behavior that undermines robustness generalization performance. Current research mainly targets identifying or mitigating a single shortcut; however, in real-world scenarios, cues within the data are diverse and unknown. In empirical studies, we reveal that the models rely to varying extents on different shortcuts. Compared to weak shortcuts, models depend more heavily on strong shortcuts, resulting in their poor generalization ability. To address these challenges, we propose MiMu, a novel method integrated with Transformer-based ERMs designed to Mitigate Multiple shortcut learning behavior, which incorporates self-calibration strategy and self-improvement strategy. In the source model, we preliminarily propose the self-calibration strategy to prevent the model from relying on shortcuts and make overconfident predictions. Then, we further design self-improvement strategy in target model to reduce the reliance on multiple shortcuts. The random mask strategy involves randomly masking partial attention positions to diversify the focus of target model other than concentrating on a fixed region. Meanwhile, the adaptive attention alignment module facilitates the alignment of attention weights to the calibrated source model, without the need for post-hoc attention maps or supervision. Finally, extensive experiments conducted on Natural Language Processing (NLP) and Computer Vision (CV) demonstrate the effectiveness of MiMu in improving robustness generalization abilities.

Mobina Pournemat, Keivan Rezaei, Gaurang Sriramanan et al.

Despite widespread success in language understanding and generation, large language models (LLMs) exhibit unclear and often inconsistent behavior when faced with tasks that require probabilistic reasoning. In this work, we present the first comprehensive study of the reasoning capabilities of LLMs over explicit discrete probability distributions. Given observations from a probability distribution, we evaluate models on three carefully designed tasks, mode identification, maximum likelihood estimation, and sample generation, by prompting them to provide responses to queries about either the joint distribution or its conditionals. These tasks thus probe a range of probabilistic skills, including frequency analysis, marginalization, and generative behavior. Through comprehensive empirical evaluations, we demonstrate that there exists a clear performance gap between smaller and larger models, with the latter demonstrating stronger inference and surprising capabilities in sample generation. Furthermore, our investigations reveal notable limitations, including sensitivity to variations in the notation utilized to represent probabilistic outcomes and performance degradation of over 60% as context length increases. Together, our results provide a detailed understanding of the probabilistic reasoning abilities of LLMs and identify key directions for future improvement.

Yuling Jiao, Yanming Lai, Defeng Sun et al.

We explore the approximation capabilities of Transformer networks for Hölder and Sobolev functions, and apply these results to address nonparametric regression estimation with dependent observations. First, we establish novel upper bounds for standard Transformer networks approximating sequence-to-sequence mappings whose component functions are Hölder continuous with smoothness index $γ\in (0,1]$. To achieve an approximation error $\varepsilon$ under the $L^p$-norm for $p \in [1, \infty]$, it suffices to use a fixed-depth Transformer network whose total number of parameters scales as $\varepsilon^{-d_x n / γ}$. This result not only extends existing findings to include the case $p = \infty$, but also matches the best known upper bounds on number of parameters previously obtained for fixed-depth FNNs and RNNs. Similar bounds are also derived for Sobolev functions. Second, we derive explicit convergence rates for the nonparametric regression problem under various $β$-mixing data assumptions, which allow the dependence between observations to weaken over time. Our bounds on the sample complexity impose no constraints on weight magnitudes. Lastly, we propose a novel proof strategy to establish approximation bounds, inspired by the Kolmogorov-Arnold representation theorem. We show that if the self-attention layer in a Transformer can perform column averaging, the network can approximate sequence-to-sequence Hölder functions, offering new insights into the interpretability of self-attention mechanisms.

Hyunin Lee, Yong Zhang, Hoang Vu Nguyen et al.

Cross-domain sequential recommendation (CDSR) aims to align heterogeneous user behavior sequences collected from different domains. While cross-attention is widely used to enhance alignment and improve recommendation performance, its underlying mechanism is not fully understood. Most researchers interpret cross-attention as residual alignment, where the output is generated by removing redundant and preserving non-redundant information from the query input by referencing another domain data which is input key and value. Beyond the prevailing view, we introduce Orthogonal Alignment, a phenomenon in which cross-attention discovers novel information that is not present in the query input, and further argue that those two contrasting alignment mechanisms can co-exist in recommendation models We find that when the query input and output of cross-attention are orthogonal, model performance improves over 300 experiments. Notably, Orthogonal Alignment emerges naturally, without any explicit orthogonality constraints. Our key insight is that Orthogonal Alignment emerges naturally because it improves scaling law. We show that baselines additionally incorporating cross-attention module outperform parameter-matched baselines, achieving a superior accuracy-per-model parameter. We hope these findings offer new directions for parameter-efficient scaling in multi-modal research.

Hamid Eghbalzadeh, Yang Wang, Rui Li et al.

Industrial ads ranking systems conventionally rely on labeled impression data, which leads to challenges such as overfitting, slower incremental gain from model scaling, and biases due to discrepancies between training and serving data. To overcome these issues, we propose a Unified framework for Knowledge-Distillation and Semi-supervised Learning (UKDSL) for ads ranking, empowering the training of models on a significantly larger and more diverse datasets, thereby reducing overfitting and mitigating training-serving data discrepancies. We provide detailed formal analysis and numerical simulations on the inherent miscalibration and prediction bias of multi-stage ranking systems, and show empirical evidence of the proposed framework's capability to mitigate those. Compared to prior work, UKDSL can enable models to learn from a much larger set of unlabeled data, hence, improving the performance while being computationally efficient. Finally, we report the successful deployment of UKDSL in an industrial setting across various ranking models, serving users at multi-billion scale, across various surfaces, geological locations, clients, and optimize for various events, which to the best of our knowledge is the first of its kind in terms of the scale and efficiency at which it operates.

Zhengyang Zhou, Qihe Huang, Binwu Wang et al.

Spatiotemporal (ST) learning has become a crucial technique to enable smart cities and sustainable urban development. Current ST learning models capture the heterogeneity via various spatial convolution and temporal evolution blocks. However, rapid urbanization leads to fluctuating distributions in urban data and city structures over short periods, resulting in existing methods suffering generalization and data adaptation issues. Despite efforts, existing methods fail to deal with newly arrived observations and those methods with generalization capacity are limited in repeated training. Motivated by complementary learning in neuroscience, we introduce a prompt-based complementary spatiotemporal learning termed ComS2T, to empower the evolution of models for data adaptation. ComS2T partitions the neural architecture into a stable neocortex for consolidating historical memory and a dynamic hippocampus for new knowledge update. We first disentangle two disjoint structures into stable and dynamic weights, and then train spatial and temporal prompts by characterizing distribution of main observations to enable prompts adaptive to new data. This data-adaptive prompt mechanism, combined with a two-stage training process, facilitates fine-tuning of the neural architecture conditioned on prompts, thereby enabling efficient adaptation during testing. Extensive experiments validate the efficacy of ComS2T in adapting to various spatiotemporal out-of-distribution scenarios while maintaining efficient inference capabilities.

Tong Chen, Hongzhi Yin, Xiangliang Zhang et al.

As a well-established approach, factorization machine (FM) is capable of automatically learning high-order interactions among features to make predictions without the need for manual feature engineering. With the prominent development of deep neural networks (DNNs), there is a recent and ongoing trend of enhancing the expressiveness of FM-based models with DNNs. However, though better results are obtained with DNN-based FM variants, such performance gain is paid off by an enormous amount (usually millions) of excessive model parameters on top of the plain FM. Consequently, the heavy parameterization impedes the real-life practicality of those deep models, especially efficient deployment on resource-constrained IoT and edge devices. In this paper, we move beyond the traditional real space where most deep FM-based models are defined, and seek solutions from quaternion representations within the hypercomplex space. Specifically, we propose the quaternion factorization machine (QFM) and quaternion neural factorization machine (QNFM), which are two novel lightweight and memory-efficient quaternion-valued models for sparse predictive analytics. By introducing a brand new take on FM-based models with the notion of quaternion algebra, our models not only enable expressive inter-component feature interactions, but also significantly reduce the parameter size due to lower degrees of freedom in the hypercomplex Hamilton product compared with real-valued matrix multiplication. Extensive experimental results on three large-scale datasets demonstrate that QFM achieves 4.36% performance improvement over the plain FM without introducing any extra parameters, while QNFM outperforms all baselines with up to two magnitudes' parameter size reduction in comparison to state-of-the-art peer methods.

Jiafa He, Chengwei Pan, Can Yang et al.

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

Yan Zhang, Zhao Zhang, Yang Wang et al.

Nonnegative matrix factorization is usually powerful for learning the "shallow" parts-based representation, but it clearly fails to discover deep hierarchical information within both the basis and representation spaces. In this paper, we technically propose a new enriched prior based Dual-constrained Deep Semi-Supervised Coupled Factorization Network, called DS2CF-Net, for learning the hierarchical coupled representations. To ex-tract hidden deep features, DS2CF-Net is modeled as a deep-structure and geometrical structure-constrained neural network. Specifically, DS2CF-Net designs a deep coupled factorization architecture using multi-layers of linear transformations, which coupled updates the bases and new representations in each layer. To improve the discriminating ability of learned deep representations and deep coefficients, our network clearly considers enriching the supervised prior by the joint deep coefficients-regularized label prediction, and incorporates enriched prior information as additional label and structure constraints. The label constraint can enable the samples of the same label to have the same coordinate in the new feature space, while the structure constraint forces the coefficient matrices in each layer to be block-diagonal so that the enhanced prior using the self-expressive label propagation are more accurate. Our network also integrates the adaptive dual-graph learning to retain the local manifold structures of both the data manifold and feature manifold by minimizing the reconstruction errors in each layer. Extensive experiments on several real databases demonstrate that our DS2CF-Net can obtain state-of-the-art performance for representation learning and clustering.

Ximing Li, Yang Wang

Partial Multi-label Learning (PML) aims to induce the multi-label predictor from datasets with noisy supervision, where each training instance is associated with several candidate labels but only partially valid. To address the noisy issue, the existing PML methods basically recover the ground-truth labels by leveraging the ground-truth confidence of the candidate label, \ie the likelihood of a candidate label being a ground-truth one. However, they neglect the information from non-candidate labels, which potentially contributes to the ground-truth label recovery. In this paper, we propose to recover the ground-truth labels, \ie estimating the ground-truth confidences, from the label enrichment, composed of the relevance degrees of candidate labels and irrelevance degrees of non-candidate labels. Upon this observation, we further develop a novel two-stage PML method, namely \emph{\underline{P}artial \underline{M}ulti-\underline{L}abel \underline{L}earning with \underline{L}abel \underline{E}nrichment-\underline{R}ecovery} (\baby), where in the first stage, it estimates the label enrichment with unconstrained label propagation, then jointly learns the ground-truth confidence and multi-label predictor given the label enrichment. Experimental results validate that \baby outperforms the state-of-the-art PML methods.

Jinjia Peng, Yang Wang, Huibing Wang et al.

Vehicle re-identification (reID) aims at identifying vehicles across different non-overlapping cameras views. The existing methods heavily relied on well-labeled datasets for ideal performance, which inevitably causes fateful drop due to the severe domain bias between the training domain and the real-world scenes; worse still, these approaches required full annotations, which is labor-consuming. To tackle these challenges, we propose a novel progressive adaptation learning method for vehicle reID, named PAL, which infers from the abundant data without annotations. For PAL, a data adaptation module is employed for source domain, which generates the images with similar data distribution to unlabeled target domain as ``pseudo target samples''. These pseudo samples are combined with the unlabeled samples that are selected by a dynamic sampling strategy to make training faster. We further proposed a weighted label smoothing (WLS) loss, which considers the similarity between samples with different clusters to balance the confidence of pseudo labels. Comprehensive experimental results validate the advantages of PAL on both VehicleID and VeRi-776 dataset.

Yang Wang

With the development of web technology, multi-modal or multi-view data has surged as a major stream for big data, where each modal/view encodes individual property of data objects. Often, different modalities are complementary to each other. Such fact motivated a lot of research attention on fusing the multi-modal feature spaces to comprehensively characterize the data objects. Most of the existing state-of-the-art focused on how to fuse the energy or information from multi-modal spaces to deliver a superior performance over their counterparts with single modal. Recently, deep neural networks have exhibited as a powerful architecture to well capture the nonlinear distribution of high-dimensional multimedia data, so naturally does for multi-modal data. Substantial empirical studies are carried out to demonstrate its advantages that are benefited from deep multi-modal methods, which can essentially deepen the fusion from multi-modal deep feature spaces. In this paper, we provide a substantial overview of the existing state-of-the-arts on the filed of multi-modal data analytics from shallow to deep spaces. Throughout this survey, we further indicate that the critical components for this field go to collaboration, adversarial competition and fusion over multi-modal spaces. Finally, we share our viewpoints regarding some future directions on this field.

Shousheng Luo, Xue-Cheng Tai, Yang Wang

We present a novel and effective binary representation for convex shapes. We show the equivalence between the shape convexity and some properties of the associated indicator function. The proposed method has two advantages. Firstly, the representation is based on a simple inequality constraint on the binary function rather than the definition of convex shapes, which allows us to obtain efficient algorithms for various applications with convexity prior. Secondly, this method is independent of the dimension of the concerned shape. In order to show the effectiveness of the proposed representation approach, we incorporate it with a probability based model for object segmentation with convexity prior. Efficient algorithms are given to solve the proposed models using Lagrange multiplier methods and linear approximations. Various experiments are given to show the superiority of the proposed methods.

Yanyan Wei, Zhao Zhang, Yang Wang et al.

Removing the rain streaks from single image is still a challenging task, since the shapes and directions of rain streaks in the synthetic datasets are very different from real images. Although supervised deep deraining networks have obtained impressive results on synthetic datasets, they still cannot obtain satisfactory results on real images due to weak generalization of rain removal capacity, i.e., the pre-trained models usually cannot handle new shapes and directions that may lead to over-derained/under-derained results. In this paper, we propose a new semi-supervised GAN-based deraining network termed Semi-DerainGAN, which can use both synthetic and real rainy images in a uniform network using two supervised and unsupervised processes. Specifically, a semi-supervised rain streak learner termed SSRML sharing the same parameters of both processes is derived, which makes the real images contribute more rain streak information. To deliver better deraining results, we design a paired discriminator for distinguishing the real pairs from fake pairs. Note that we also contribute a new real-world rainy image dataset Real200 to alleviate the difference between the synthetic and real image do-mains. Extensive results on public datasets show that our model can obtain competitive performance, especially on real images.

Zhao Zhang, Zemin Tang, Yang Wang et al.

Deep Convolutional Neural Networks (CNNs), such as Dense Convolutional Networks (DenseNet), have achieved great success for image representation by discovering deep hierarchical information. However, most existing networks simply stacks the convolutional layers and hence failing to fully discover local and global feature information among layers. In this paper, we mainly explore how to enhance the local and global dense feature flow by exploiting hierarchical features fully from all the convolution layers. Technically, we propose an efficient and effective CNN framework, i.e., Fast Dense Residual Network (FDRN), for text recognition. To construct FDRN, we propose a new fast residual dense block (f-RDB) to retain the ability of local feature fusion and local residual learning of original RDB, which can reduce the computing efforts at the same time. After fully learning local residual dense features, we utilize the sum operation and several f-RDBs to define a new block termed global dense block (GDB) by imitating the construction of dense blocks to learn global dense residual features adaptively in a holistic way. Finally, we use two convolution layers to construct a down-sampling block to reduce the global feature size and extract deeper features. Extensive simulations show that FDRN obtains the enhanced recognition results, compared with other related models.

Jiahuan Ren, Zhao Zhang, Sheng Li et al.

In this paper, we investigate the robust dictionary learning (DL) to discover the hybrid salient low-rank and sparse representation in a factorized compressed space. A Joint Robust Factorization and Projective Dictionary Learning (J-RFDL) model is presented. The setting of J-RFDL aims at improving the data representations by enhancing the robustness to outliers and noise in data, encoding the reconstruction error more accurately and obtaining hybrid salient coefficients with accurate reconstruction ability. Specifically, J-RFDL performs the robust representation by DL in a factorized compressed space to eliminate the negative effects of noise and outliers on the results, which can also make the DL process efficient. To make the encoding process robust to noise in data, J-RFDL clearly uses sparse L2, 1-norm that can potentially minimize the factorization and reconstruction errors jointly by forcing rows of the reconstruction errors to be zeros. To deliver salient coefficients with good structures to reconstruct given data well, J-RFDL imposes the joint low-rank and sparse constraints on the embedded coefficients with a synthesis dictionary. Based on the hybrid salient coefficients, we also extend J-RFDL for the joint classification and propose a discriminative J-RFDL model, which can improve the discriminating abilities of learnt coeffi-cients by minimizing the classification error jointly. Extensive experiments on public datasets demonstrate that our formulations can deliver superior performance over other state-of-the-art methods.

Zhao Zhang, Yulin Sun, Yang Wang et al.

Both the Dictionary Learning (DL) and Convolutional Neural Networks (CNN) are powerful image representation learning systems based on different mechanisms and principles, however whether we can seamlessly integrate them to improve the per-formance is noteworthy exploring. To address this issue, we propose a novel generalized end-to-end representation learning architecture, dubbed Convolutional Dictionary Pair Learning Network (CDPL-Net) in this paper, which integrates the learning schemes of the CNN and dictionary pair learning into a unified framework. Generally, the architecture of CDPL-Net includes two convolutional/pooling layers and two dictionary pair learn-ing (DPL) layers in the representation learning module. Besides, it uses two fully-connected layers as the multi-layer perception layer in the nonlinear classification module. In particular, the DPL layer can jointly formulate the discriminative synthesis and analysis representations driven by minimizing the batch based reconstruction error over the flatted feature maps from the convolution/pooling layer. Moreover, DPL layer uses l1-norm on the analysis dictionary so that sparse representation can be delivered, and the embedding process will also be robust to noise. To speed up the training process of DPL layer, the efficient stochastic gradient descent is used. Extensive simulations on real databases show that our CDPL-Net can deliver enhanced performance over other state-of-the-art methods.

Zhao Zhang, Zemin Tang, Yang Wang et al.

Convolutional Recurrent Neural Network (CRNN) is a popular network for recognizing texts in images. Advances like the variant of CRNN, such as Dense Convolutional Network with Connectionist Temporal Classification, has reduced the running time of the network, but exposing the inner computation cost and weight size of the convolutional networks as a bottleneck. Specifically, the DenseNet based models utilize the dense blocks as the core module, but the inner features are combined in the form of concatenation in dense blocks. As such, the number of channels of combined features delivered as the input of the layers close to the output and the relevant computational cost grows rapidly with the dense blocks getting deeper. This will severely bring heavy computational cost and big weight size, which restrict the depth of dense blocks. In this paper, we propose a compressed convolution block called Lightweight Dense Block (LDB). To reduce the computing cost and weight size, we re-define and re-design the way of combining internal features of the dense blocks. LDB is a convolutional block similarly as dense block, but it can reduce the computation cost and weight size to (1/L, 2/L), compared with original ones, where L is the number of layers in blocks. Moreover, LDB can be used to replace the original dense block in any DenseNet based models. Based on the LDBs, we propose a Compressed DenseNet (CDenseNet) for the lightweight character recognition. Extensive experiments demonstrate that CDenseNet can effectively reduce the weight size while delivering the promising recognition results.

Yanyan Wei, Zhao Zhang, Yang Wang et al.

Single image deraining (SID) is an important and challenging topic in emerging vision applications, and most of emerged deraining methods are supervised relying on the ground truth (i.e., paired images) in recent years. However, in practice it is rather common to have no un-paired images in real deraining task, in such cases how to remove the rain streaks in an unsupervised way will be a very challenging task due to lack of constraints between images and hence suffering from low-quality recovery results. In this paper, we explore the unsupervised SID task using unpaired data and propose a novel net called Attention-guided Deraining by Constrained CycleGAN (or shortly, DerainCycleGAN), which can fully utilize the constrained transfer learning abilitiy and circulatory structure of CycleGAN. Specifically, we design an unsu-pervised attention guided rain streak extractor (U-ARSE) that utilizes a memory to extract the rain streak masks with two constrained cycle-consistency branches jointly by paying attention to both the rainy and rain-free image domains. As a by-product, we also contribute a new paired rain image dataset called Rain200A, which is constructed by our network automatically. Compared with existing synthesis datasets, the rainy streaks in Rain200A contains more obvious and diverse shapes and directions. As a result, existing supervised methods trained on Rain200A can perform much better for processing real rainy images. Extensive experiments on synthesis and real datasets show that our net is superior to existing unsupervised deraining networks, and is also very competitive to other related supervised networks.

Yang Wang, Yang Cao, Zheng-Jun Zha et al.

Contrast enhancement and noise removal are coupled problems for low-light image enhancement. The existing Retinex based methods do not take the coupling relation into consideration, resulting in under or over-smoothing of the enhanced images. To address this issue, this paper presents a novel progressive Retinex framework, in which illumination and noise of low-light image are perceived in a mutually reinforced manner, leading to noise reduction low-light enhancement results. Specifically, two fully pointwise convolutional neural networks are devised to model the statistical regularities of ambient light and image noise respectively, and to leverage them as constraints to facilitate the mutual learning process. The proposed method not only suppresses the interference caused by the ambiguity between tiny textures and image noises, but also greatly improves the computational efficiency. Moreover, to solve the problem of insufficient training data, we propose an image synthesis strategy based on camera imaging model, which generates color images corrupted by illumination-dependent noises. Experimental results on both synthetic and real low-light images demonstrate the superiority of our proposed approaches against the State-Of-The-Art (SOTA) low-light enhancement methods.

Biao Qian, Yang Wang

1 bit deep neural networks (DNNs), of which both the activations and weights are binarized , are attracting more and more attention due to their high computational efficiency and low memory requirement . However, the drawback of large accuracy dropping also restrict s its application. In this paper, we propose a novel Targeted Acceleration and Compression (TAC) framework to improve the performance of 1 bit deep neural networks W e consider that the acceleration and compression effects of binarizing fully connected layer s are not sufficient to compensate for the accuracy loss caused by it In the proposed framework, t he convolutional and fully connected layer are separated and optimized i ndividually . F or the convolutional layer s , both the activations and weights are binarized. For the fully connected layer s, the binarization operation is re placed by network pruning and low bit quantization. The proposed framework is implemented on the CIFAR 10, CIFAR 100 and ImageNet ( ILSVRC 12 ) datasets , and experimental results show that the proposed TAC can significantly improve the accuracy of 1 bit deep neural networks and outperforms the state of the art by more than 6 percentage points .

Linwei Ye, Mrigank Rochan, Zhi Liu et al.

We consider the problem of referring image segmentation. Given an input image and a natural language expression, the goal is to segment the object referred by the language expression in the image. Existing works in this area treat the language expression and the input image separately in their representations. They do not sufficiently capture long-range correlations between these two modalities. In this paper, we propose a cross-modal self-attention (CMSA) module that effectively captures the long-range dependencies between linguistic and visual features. Our model can adaptively focus on informative words in the referring expression and important regions in the input image. In addition, we propose a gated multi-level fusion module to selectively integrate self-attentive cross-modal features corresponding to different levels in the image. This module controls the information flow of features at different levels. We validate the proposed approach on four evaluation datasets. Our proposed approach consistently outperforms existing state-of-the-art methods.

Tanzila Rahman, Mrigank Rochan, Yang Wang

The goal of video-based person re-identification is to match two input videos, so that the distance of the two videos is small if two videos contain the same person. A common approach for person re-identification is to first extract image features for all frames in the video, then aggregate all the features to form a video-level feature. The video-level features of two videos can then be used to calculate the distance of the two videos. In this paper, we propose a temporal attention approach for aggregating frame-level features into a video-level feature vector for re-identification. Our method is motivated by the fact that not all frames in a video are equally informative. We propose a fully convolutional temporal attention model for generating the attention scores. Fully convolutional network (FCN) has been widely used in semantic segmentation for generating 2D output maps. In this paper, we formulate video based person re-identification as a sequence labeling problem like semantic segmentation. We establish a connection between them and modify FCN to generate attention scores to represent the importance of each frame. Extensive experiments on three different benchmark datasets (i.e. iLIDS-VID, PRID-2011 and SDU-VID) show that our proposed method outperforms other state-of-the-art approaches.

Lin Wu, Richang Hong, Yang Wang et al.

Person re-identification (re-ID) is a task of matching pedestrians under disjoint camera views. To recognise paired snapshots, it has to cope with large cross-view variations caused by the camera view shift. Supervised deep neural networks are effective in producing a set of non-linear projections that can transform cross-view images into a common feature space. However, they typically impose a symmetric architecture, yielding the network ill-conditioned on its optimisation. In this paper, we learn view-invariant subspace for person re-ID, and its corresponding similarity metric using an adversarial view adaptation approach. The main contribution is to learn coupled asymmetric mappings regarding view characteristics which are adversarially trained to address the view discrepancy by optimising the cross-entropy view confusion objective. To determine the similarity value, the network is empowered with a similarity discriminator to promote features that are highly discriminant in distinguishing positive and negative pairs. The other contribution includes an adaptive weighing on the most difficult samples to address the imbalance of within/between-identity pairs. Our approach achieves notable improved performance in comparison to state-of-the-arts on benchmark datasets.

Lin Wu, Yang Wang, Hongzhi Yin et al.

Video-based person re-identification (re-ID) refers to matching people across camera views from arbitrary unaligned video footages. Existing methods rely on supervision signals to optimise a projected space under which the distances between inter/intra-videos are maximised/minimised. However, this demands exhaustively labelling people across camera views, rendering them unable to be scaled in large networked cameras. Also, it is noticed that learning effective video representations with view invariance is not explicitly addressed for which features exhibit different distributions otherwise. Thus, matching videos for person re-ID demands flexible models to capture the dynamics in time-series observations and learn view-invariant representations with access to limited labeled training samples. In this paper, we propose a novel few-shot deep learning approach to video-based person re-ID, to learn comparable representations that are discriminative and view-invariant. The proposed method is developed on the variational recurrent neural networks (VRNNs) and trained adversarially to produce latent variables with temporal dependencies that are highly discriminative yet view-invariant in matching persons. Through extensive experiments conducted on three benchmark datasets, we empirically show the capability of our method in creating view-invariant temporal features and state-of-the-art performance achieved by our method.

Lin Wu, Yang Wang, Ling Shao et al.

In this paper, we introduce a global video representation to video-based person re-identification (re-ID) that aggregates local 3D features across the entire video extent. Most of the existing methods rely on 2D convolutional networks (ConvNets) to extract frame-wise deep features which are pooled temporally to generate the video-level representations. However, 2D ConvNets lose temporal input information immediately after the convolution, and a separate temporal pooling is limited in capturing human motion in shorter sequences. To this end, we present a \textit{global} video representation (3D PersonVLAD), complementary to 3D ConvNets as a novel layer to capture the appearance and motion dynamics in full-length videos. However, encoding each video frame in its entirety and computing an aggregate global representation across all frames is tremendously challenging due to occlusions and misalignments. To resolve this, our proposed network is further augmented with 3D part alignment module to learn local features through soft-attention module. These attended features are statistically aggregated to yield identity-discriminative representations. Our global 3D features are demonstrated to achieve state-of-the-art results on three benchmark datasets: MARS \cite{MARS}, iLIDS-VID \cite{VideoRanking}, and PRID 2011

Shenglan Liu, Shuai Guo, Hong Qiao et al.

Human emotion recognition is an important direction in the field of biometric and information forensics. However, most existing human emotion research are based on the single RGB view. In this paper, we introduce a RGBD video-emotion dataset and a RGBD face-emotion dataset for research. To our best knowledge, this may be the first RGBD video-emotion dataset. We propose a new supervised nonlinear multi-view laplacian eigenmaps (MvLE) approach and a multihidden-layer out-of-sample network (MHON) for RGB-D humanemotion recognition. To get better representations of RGB view and depth view, MvLE is used to map the training set of both views from original space into the common subspace. As RGB view and depth view lie in different spaces, a new distance metric bag of neighbors (BON) used in MvLE can get the similar distributions of the two views. Finally, MHON is used to get the low-dimensional representations of test data and predict their labels. MvLE can deal with the cases that RGB view and depth view have different size of features, even different number of samples and classes. And our methods can be easily extended to more than two views. The experiment results indicate the effectiveness of our methods over some state-of-art methods.

Shivansh Rao, Tanzila Rahman, Mrigank Rochan et al.

We consider the problem of video-based person re-identification. The goal is to identify a person from videos captured under different cameras. In this paper, we propose an efficient spatial-temporal attention based model for person re-identification from videos. Our method generates an attention score for each frame based on frame-level features. The attention scores of all frames in a video are used to produce a weighted feature vector for the input video. Unlike most existing deep learning methods that use global representation, our approach focuses on attention scores. Extensive experiments on two benchmark datasets demonstrate that our method achieves the state-of-the-art performance. This is a technical report.

Shenglan Liu, Xiang Liu, Yang Liu et al.

Supervised learning methods are widely used in machine learning. However, the lack of labels in existing data limits the application of these technologies. Visual interactive learning (VIL) compared with computers can avoid semantic gap, and solve the labeling problem of small label quantity (SLQ) samples in a groundbreaking way. In order to fully understand the importance of VIL to the interaction process, we re-summarize the interactive learning related algorithms (e.g. clustering, classification, retrieval etc.) from the perspective of VIL. Note that, perception and cognition are two main visual processes of VIL. On this basis, we propose a perceptual visual interactive learning (PVIL) framework, which adopts gestalt principle to design interaction strategy and multi-dimensionality reduction (MDR) to optimize the process of visualization. The advantage of PVIL framework is that it combines computer's sensitivity of detailed features and human's overall understanding of global tasks. Experimental results validate that the framework is superior to traditional computer labeling methods (such as label propagation) in both accuracy and efficiency, which achieves significant classification results on dense distribution and sparse classes dataset.

Lin Wu, Yang Wang, Junbin Gao et al.

Person re-identification (re-ID) requires rapid, flexible yet discriminant representations to quickly generalize to unseen observations on-the-fly and recognize the same identity across disjoint camera views. Recent effective methods are developed in a pair-wise similarity learning system to detect a fixed set of features from distinct regions which are mapped to their vector embeddings for the distance measuring. However, the most relevant and crucial parts of each image are detected independently without referring to the dependency conditioned on one and another. Also, these region based methods rely on spatial manipulation to position the local features in comparable similarity measuring. To combat these limitations, in this paper we introduce the Deep Co-attention based Comparators (DCCs) that fuse the co-dependent representations of the paired images so as to focus on the relevant parts of both images and produce their \textit{relative representations}. Given a pair of pedestrian images to be compared, the proposed model mimics the foveation of human eyes to detect distinct regions concurrent on both images, namely co-dependent features, and alternatively attend to relevant regions to fuse them into the similarity learning. Our comparator is capable of producing dynamic representations relative to a particular sample every time, and thus well-suited to the case of re-identifying pedestrians on-the-fly. We perform extensive experiments to provide the insights and demonstrate the effectiveness of the proposed DCCs in person re-ID. Moreover, our approach has achieved the state-of-the-art performance on three benchmark data sets: DukeMTMC-reID \cite{DukeMTMC}, CUHK03 \cite{FPNN}, and Market-1501 \cite{Market1501}.

Zhenheng Yang, Peng Wang, Yang Wang et al.

Learning to estimate 3D geometry in a single image by watching unlabeled videos via deep convolutional network is attracting significant attention. In this paper, we introduce a "3D as-smooth-as-possible (3D-ASAP)" prior inside the pipeline, which enables joint estimation of edges and 3D scene, yielding results with significant improvement in accuracy for fine detailed structures. Specifically, we define the 3D-ASAP prior by requiring that any two points recovered in 3D from an image should lie on an existing planar surface if no other cues provided. We design an unsupervised framework that Learns Edges and Geometry (depth, normal) all at Once (LEGO). The predicted edges are embedded into depth and surface normal smoothness terms, where pixels without edges in-between are constrained to satisfy the prior. In our framework, the predicted depths, normals and edges are forced to be consistent all the time. We conduct experiments on KITTI to evaluate our estimated geometry and CityScapes to perform edge evaluation. We show that in all of the tasks, i.e.depth, normal and edge, our algorithm vastly outperforms other state-of-the-art (SOTA) algorithms, demonstrating the benefits of our approach.

Jing Zhang, Yang Cao, Yang Wang et al.

Modeling statistical regularity plays an essential role in ill-posed image processing problems. Recently, deep learning based methods have been presented to implicitly learn statistical representation of pixel distributions in natural images and leverage it as a constraint to facilitate subsequent tasks, such as color constancy and image dehazing. However, the existing CNN architecture is prone to variability and diversity of pixel intensity within and between local regions, which may result in inaccurate statistical representation. To address this problem, this paper presents a novel fully point-wise CNN architecture for modeling statistical regularities in natural images. Specifically, we propose to randomly shuffle the pixels in the origin images and leverage the shuffled image as input to make CNN more concerned with the statistical properties. Moreover, since the pixels in the shuffled image are independent identically distributed, we can replace all the large convolution kernels in CNN with point-wise ($1*1$) convolution kernels while maintaining the representation ability. Experimental results on two applications: color constancy and image dehazing, demonstrate the superiority of our proposed network over the existing architectures, i.e., using 1/10$\sim$1/100 network parameters and computational cost while achieving comparable performance.

Yang Wang, Yi Yang, Zhenheng Yang et al.

It has been recently shown that a convolutional neural network can learn optical flow estimation with unsupervised learning. However, the performance of the unsupervised methods still has a relatively large gap compared to its supervised counterpart. Occlusion and large motion are some of the major factors that limit the current unsupervised learning of optical flow methods. In this work we introduce a new method which models occlusion explicitly and a new warping way that facilitates the learning of large motion. Our method shows promising results on Flying Chairs, MPI-Sintel and KITTI benchmark datasets. Especially on KITTI dataset where abundant unlabeled samples exist, our unsupervised method outperforms its counterpart trained with supervised learning.

Lin Wu, Yang Wang

Fine-grained visual recognition typically depends on modeling subtle difference from object parts. However, these parts often exhibit dramatic visual variations such as occlusions, viewpoints, and spatial transformations, making it hard to detect. In this paper, we present a novel attention-based model to automatically, selectively and accurately focus on critical object regions with higher importance against appearance variations. Given an image, two different Convolutional Neural Networks (CNNs) are constructed, where the outputs of two CNNs are correlated through bilinear pooling to simultaneously focus on discriminative regions and extract relevant features. To capture spatial distributions among the local regions with visual attention, soft attention based spatial Long-Short Term Memory units (LSTMs) are incorporated to realize spatially recurrent yet visually selective over local input patterns. All the above intuitions equip our network with the following novel model: two-stream CNN layers, bilinear pooling layer, spatial recurrent layer with location attention are jointly trained via an end-to-end fashion to serve as the part detector and feature extractor, whereby relevant features are localized and extracted attentively. We show the significance of our network against two well-known visual recognition tasks: fine-grained image classification and person re-identification.

Yang Wang, Lin Wu

Multi-view data clustering attracts more attention than their single view counterparts due to the fact that leveraging multiple independent and complementary information from multi-view feature spaces outperforms the single one. Multi-view Spectral Clustering aims at yielding the data partition agreement over their local manifold structures by seeking eigenvalue-eigenvector decompositions. However, as we observed, such classical paradigm still suffers from (1) overlooking the flexible local manifold structure, caused by (2) enforcing the low-rank data correlation agreement among all views; worse still, (3) LRR is not intuitively flexible to capture the latent data clustering structures. In this paper, we present the structured LRR by factorizing into the latent low-dimensional data-cluster representations, which characterize the data clustering structure for each view. Upon such representation, (b) the laplacian regularizer is imposed to be capable of preserving the flexible local manifold structure for each view. (c) We present an iterative multi-view agreement strategy by minimizing the divergence objective among all factorized latent data-cluster representations during each iteration of optimization process, where such latent representation from each view serves to regulate those from other views, such intuitive process iteratively coordinates all views to be agreeable. (d) We remark that such data-cluster representation can flexibly encode the data clustering structure from any view with adaptive input cluster number. To this end, (e) a novel non-convex objective function is proposed via the efficient alternating minimization strategy. The complexity analysis are also presented. The extensive experiments conducted against the real-world multi-view datasets demonstrate the superiority over state-of-the-arts.

Yang Wang, Lin Wu

Low-Rank Representation (LRR) is arguably one of the most powerful paradigms for Multi-view spectral clustering, which elegantly encodes the multi-view local graph/manifold structures into an intrinsic low-rank self-expressive data similarity embedded in high-dimensional space, to yield a better graph partition than their single-view counterparts. In this paper we revisit it with a fundamentally different perspective by discovering LRR as essentially a latent clustered orthogonal projection based representation winged with an optimized local graph structure for spectral clustering; each column of the representation is fundamentally a cluster basis orthogonal to others to indicate its members, which intuitively projects the view-specific feature representation to be the one spanned by all orthogonal basis to characterize the cluster structures. Upon this finding, we propose our technique with the followings: (1) We decompose LRR into latent clustered orthogonal representation via low-rank matrix factorization, to encode the more flexible cluster structures than LRR over primal data objects; (2) We convert the problem of LRR into that of simultaneously learning orthogonal clustered representation and optimized local graph structure for each view; (3) The learned orthogonal clustered representations and local graph structures enjoy the same magnitude for multi-view, so that the ideal multi-view consensus can be readily achieved. The experiments over multi-view datasets validate its superiority.

Lin Wu, Yang Wang, Xue Li et al.

Matching pedestrians across disjoint camera views, known as person re-identification (re-id), is a challenging problem that is of importance to visual recognition and surveillance. Most existing methods exploit local regions within spatial manipulation to perform matching in local correspondence. However, they essentially extract \emph{fixed} representations from pre-divided regions for each image and perform matching based on the extracted representation subsequently. For models in this pipeline, local finer patterns that are crucial to distinguish positive pairs from negative ones cannot be captured, and thus making them underperformed. In this paper, we propose a novel deep multiplicative integration gating function, which answers the question of \emph{what-and-where to match} for effective person re-id. To address \emph{what} to match, our deep network emphasizes common local patterns by learning joint representations in a multiplicative way. The network comprises two Convolutional Neural Networks (CNNs) to extract convolutional activations, and generates relevant descriptors for pedestrian matching. This thus, leads to flexible representations for pair-wise images. To address \emph{where} to match, we combat the spatial misalignment by performing spatially recurrent pooling via a four-directional recurrent neural network to impose spatial dependency over all positions with respect to the entire image. The proposed network is designed to be end-to-end trainable to characterize local pairwise feature interactions in a spatially aligned manner. To demonstrate the superiority of our method, extensive experiments are conducted over three benchmark data sets: VIPeR, CUHK03 and Market-1501.

Lin Wu, Yang Wang, Junbin Gao et al.

Person re-identification (re-id) aims to match pedestrians observed by disjoint camera views. It attracts increasing attention in computer vision due to its importance to surveillance system. To combat the major challenge of cross-view visual variations, deep embedding approaches are proposed by learning a compact feature space from images such that the Euclidean distances correspond to their cross-view similarity metric. However, the global Euclidean distance cannot faithfully characterize the ideal similarity in a complex visual feature space because features of pedestrian images exhibit unknown distributions due to large variations in poses, illumination and occlusion. Moreover, intra-personal training samples within a local range are robust to guide deep embedding against uncontrolled variations, which however, cannot be captured by a global Euclidean distance. In this paper, we study the problem of person re-id by proposing a novel sampling to mine suitable \textit{positives} (i.e. intra-class) within a local range to improve the deep embedding in the context of large intra-class variations. Our method is capable of learning a deep similarity metric adaptive to local sample structure by minimizing each sample's local distances while propagating through the relationship between samples to attain the whole intra-class minimization. To this end, a novel objective function is proposed to jointly optimize similarity metric learning, local positive mining and robust deep embedding. This yields local discriminations by selecting local-ranged positive samples, and the learned features are robust to dramatic intra-class variations. Experiments on benchmarks show state-of-the-art results achieved by our method.

Lin Wu, Yang Wang

Given a pedestrian image as a query, the purpose of person re-identification is to identify the correct match from a large collection of gallery images depicting the same person captured by disjoint camera views. The critical challenge is how to construct a robust yet discriminative feature representation to capture the compounded variations in pedestrian appearance. To this end, deep learning methods have been proposed to extract hierarchical features against extreme variability of appearance. However, existing methods in this category generally neglect the efficiency in the matching stage whereas the searching speed of a re-identification system is crucial in real-world applications. In this paper, we present a novel deep hashing framework with Convolutional Neural Networks (CNNs) for fast person re-identification. Technically, we simultaneously learn both CNN features and hash functions/codes to get robust yet discriminative features and similarity-preserving hash codes. Thereby, person re-identification can be resolved by efficiently computing and ranking the Hamming distances between images. A structured loss function defined over positive pairs and hard negatives is proposed to formulate a novel optimization problem so that fast convergence and more stable optimized solution can be obtained. Extensive experiments on two benchmarks CUHK03 \cite{FPNN} and Market-1501 \cite{Market1501} show that the proposed deep architecture is efficacy over state-of-the-arts.

Yang Wang, Vinh Tran, Minh Hoai

Recently Trajectory-pooled Deep-learning Descriptors were shown to achieve state-of-the-art human action recognition results on a number of datasets. This paper improves their performance by applying rank pooling to each trajectory, encoding the temporal evolution of deep learning features computed along the trajectory. This leads to Evolution-Preserving Trajectory (EPT) descriptors, a novel type of video descriptor that significantly outperforms Trajectory-pooled Deep-learning Descriptors. EPT descriptors are defined based on dense trajectories, and they provide complimentary benefits to video descriptors that are not based on trajectories. In particular, we show that the combination of EPT descriptors and VideoDarwin leads to state-of-the-art performance on Hollywood2 and UCF101 datasets.

Yang Wang, Xuemin Lin, Lin Wu et al.

Given a query photo issued by a user (q-user), the landmark retrieval is to return a set of photos with their landmarks similar to those of the query, while the existing studies on the landmark retrieval focus on exploiting geometries of landmarks for similarity matches between candidate photos and a query photo. We observe that the same landmarks provided by different users over social media community may convey different geometry information depending on the viewpoints and/or angles, and may subsequently yield very different results. In fact, dealing with the landmarks with \illshapes caused by the photography of q-users is often nontrivial and has seldom been studied. In this paper we propose a novel framework, namely multi-query expansions, to retrieve semantically robust landmarks by two steps. Firstly, we identify the top-$k$ photos regarding the latent topics of a query landmark to construct multi-query set so as to remedy its possible \illshape. For this purpose, we significantly extend the techniques of Latent Dirichlet Allocation. Then, motivated by the typical \emph{collaborative filtering} methods, we propose to learn a \emph{collaborative} deep networks based semantically, nonlinear and high-level features over the latent factor for landmark photo as the training set, which is formed by matrix factorization over \emph{collaborative} user-photo matrix regarding the multi-query set. The learned deep network is further applied to generate the features for all the other photos, meanwhile resulting into a compact multi-query set within such space. Extensive experiments are conducted on real-world social media data with both landmark photos together with their user information to show the superior performance over the existing methods.

Xingzhong Du, Hongzhi Yin, Ling Chen et al.

Video recommendation has become an essential way of helping people explore the massive videos and discover the ones that may be of interest to them. In the existing video recommender systems, the models make the recommendations based on the user-video interactions and single specific content features. When the specific content features are unavailable, the performance of the existing models will seriously deteriorate. Inspired by the fact that rich contents (e.g., text, audio, motion, and so on) exist in videos, in this paper, we explore how to use these rich contents to overcome the limitations caused by the unavailability of the specific ones. Specifically, we propose a novel general framework that incorporates arbitrary single content feature with user-video interactions, named as collaborative embedding regression (CER) model, to make effective video recommendation in both in-matrix and out-of-matrix scenarios. Our extensive experiments on two real-world large-scale datasets show that CER beats the existing recommender models with any single content feature and is more time efficient. In addition, we propose a priority-based late fusion (PRI) method to gain the benefit brought by the integrating the multiple content features. The corresponding experiment shows that PRI brings real performance improvement to the baseline and outperforms the existing fusion methods.

Lin Wu, Yang Wang

Learning hash functions/codes for similarity search over multi-view data is attracting increasing attention, where similar hash codes are assigned to the data objects characterizing consistently neighborhood relationship across views. Traditional methods in this category inherently suffer three limitations: 1) they commonly adopt a two-stage scheme where similarity matrix is first constructed, followed by a subsequent hash function learning; 2) these methods are commonly developed on the assumption that data samples with multiple representations are noise-free,which is not practical in real-life applications; 3) they often incur cumbersome training model caused by the neighborhood graph construction using all $N$ points in the database ($O(N)$). In this paper, we motivate the problem of jointly and efficiently training the robust hash functions over data objects with multi-feature representations which may be noise corrupted. To achieve both the robustness and training efficiency, we propose an approach to effectively and efficiently learning low-rank kernelized \footnote{We use kernelized similarity rather than kernel, as it is not a squared symmetric matrix for data-landmark affinity matrix.} hash functions shared across views. Specifically, we utilize landmark graphs to construct tractable similarity matrices in multi-views to automatically discover neighborhood structure in the data. To learn robust hash functions, a latent low-rank kernel function is used to construct hash functions in order to accommodate linearly inseparable data. In particular, a latent kernelized similarity matrix is recovered by rank minimization on multiple kernel-based similarity matrices. Extensive experiments on real-world multi-view datasets validate the efficacy of our method in the presence of error corruptions.