Classifying the confusing samples in the course of RGBT tracking is a quite challenging problem, which hasn't got satisfied solution. Existing methods only focus on enlarging the boundary between positive and negative samples, however, the structured information of samples might be harmed, e.g., confusing positive samples are closer to the anchor than normal positive samples.To handle this problem, we propose a novel Multi-Modal Multi-Margin Metric Learning framework, named M$^5$L for RGBT tracking in this paper. In particular, we design a multi-margin structured loss to distinguish the confusing samples which play a most critical role in tracking performance boosting. To alleviate this problem, we additionally enlarge the boundaries between confusing positive samples and normal ones, between confusing negative samples and normal ones with predefined margins, by exploiting the structured information of all samples in each modality.Moreover, a cross-modality constraint is employed to reduce the difference between modalities and push positive samples closer to the anchor than negative ones from two modalities.In addition, to achieve quality-aware RGB and thermal feature fusion, we introduce the modality attentions and learn them using a feature fusion module in our network. Extensive experiments on large-scale datasets testify that our framework clearly improves the tracking performance and outperforms the state-of-the-art RGBT trackers.
Person search is to detect all persons and identify the query persons from detected persons in the image without proposals and bounding boxes, which is different from person re-identification. In this paper, we propose a fusing multi-task convolutional neural network(FMT-CNN) to tackle the correlation and heterogeneity of detection and re-identification with a single convolutional neural network. We focus on how the interplay of person detection and person re-identification affects the overall performance. We employ person labels in region proposal network to produce features for person re-identification and person detection network, which can improve the accuracy of detection and re-identification simultaneously. We also use a multiple loss to train our re-identification network. Experiment results on CUHK-SYSU Person Search dataset show that the performance of our proposed method is superior to state-of-the-art approaches in both mAP and top-1.
Image salient object detection (SOD) is an active research topic in computer vision and multimedia area. Fusing complementary information of RGB and depth has been demonstrated to be effective for image salient object detection which is known as RGB-D salient object detection problem. The main challenge for RGB-D salient object detection is how to exploit the salient cues of both intra-modality (RGB, depth) and cross-modality simultaneously which is known as cross-modality detection problem. In this paper, we tackle this challenge by designing a novel cross-modality Saliency Generative Adversarial Network (\emph{cm}SalGAN). \emph{cm}SalGAN aims to learn an optimal view-invariant and consistent pixel-level representation for RGB and depth images via a novel adversarial learning framework, which thus incorporates both information of intra-view and correlation information of cross-view images simultaneously for RGB-D saliency detection problem. To further improve the detection results, the attention mechanism and edge detection module are also incorporated into \emph{cm}SalGAN. The entire \emph{cm}SalGAN can be trained in an end-to-end manner by using the standard deep neural network framework. Experimental results show that \emph{cm}SalGAN achieves the new state-of-the-art RGB-D saliency detection performance on several benchmark datasets.
Discriminative feature representation of person image is important for person re-identification (Re-ID) task. Recently, attributes have been demonstrated beneficially in guiding for learning more discriminative feature representations for Re-ID. As attributes normally co-occur in person images, it is desirable to model the attribute dependencies to improve the attribute prediction and thus Re-ID results. In this paper, we propose to model these attribute dependencies via a novel attribute knowledge graph (AttKG), and propose a novel Attribute Knowledge Graph Convolutional Network (AttKGCN) to solve Re-ID problem. AttKGCN integrates both attribute prediction and Re-ID learning together in a unified end-to-end framework which can boost their performances, respectively. AttKGCN first builds a directed attribute KG whose nodes denote attributes and edges encode the co-occurrence relationships of different attributes. Then, AttKGCN learns a set of inter-dependent attribute classifiers which are combined with person visual descriptors for attribute prediction. Finally, AttKGCN integrates attribute description and deeply visual representation together to construct a more discriminative feature representation for Re-ID task. Extensive experiments on several benchmark datasets demonstrate the effectiveness of AttKGCN on attribute prediction and Re-ID tasks.
Recently, graph convolutional networks (GCNs) have shown great potential for the task of graph matching. It can integrate graph node feature embedding, node-wise affinity learning and matching optimization together in a unified end-to-end model. One important aspect of graph matching is the construction of two matching graphs. However, the matching graphs we feed to existing graph convolutional matching networks are generally fixed and independent of graph matching, which thus are not guaranteed to be optimal for the graph matching task. Also, existing GCN matching method employs several general smoothing-based graph convolutional layers to generate graph node embeddings, in which extensive smoothing convolution operation may dilute the desired discriminatory information of graph nodes. To overcome these issues, we propose a novel Graph Learning-Matching Network (GLMNet) for graph matching problem. GLMNet has three main aspects. (1) It integrates graph learning into graph matching which thus adaptively learn a pair of optimal graphs that best serve graph matching task. (2) It further employs a Laplacian sharpening convolutional module to generate more discriminative node embeddings for graph matching. (3) A new constraint regularized loss is designed for GLMNet training which can encode the desired one-to-one matching constraints in matching optimization. Experiments on two benchmarks demonstrate the effectiveness of GLMNet and advantages of its main modules.
Pose prediction is an increasingly interesting topic in computer vision and robotics. In this paper, we propose a new network, Trajectorylet-Net, to predict future poses. Compared with most existing methods, our model focuses on modeling the co-occurrence long-term information and spatiotemporal correlation. Specifically, a novel descriptor, trajectorylet, is introduced to characterize the static and dynamic information of the input pose sequence. Then, a coupled spatio-temporal learning schema is proposed to generate trajectorylet descriptors, which can simultaneously capture the local structure of the human body and the global co-occurrence temporal information of the input sequence. Finally, we propose to predict future poses by gathering trajectorylet descriptors gradually. Extensive experiments show that our method achieves state-of-the-art performance on two benchmarks (e.g. G3D and FNTU), which demonstrates the effectiveness of our proposed method.
Graph Neural Networks (GNNs) have been widely studied for graph data representation and learning. However, existing GNNs generally conduct context-aware learning on node feature representation only which usually ignores the learning of edge (weight) representation. In this paper, we propose a novel unified GNN model, named Context-aware Adaptive Graph Attention Network (CaGAT). CaGAT aims to learn a context-aware attention representation for each graph edge by further exploiting the context relationships among different edges. In particular, CaGAT conducts context-aware learning on both node feature representation and edge (weight) representation simultaneously and cooperatively in a unified manner which can boost their respective performance in network training. We apply CaGAT on semi-supervised learning tasks. Promising experimental results on several benchmark datasets demonstrate the effectiveness and benefits of CaGAT.
Graph Convolutional Networks (GCNs) have been widely studied for graph data representation and learning. The graph convolution operation in GCNs can usually be regarded as a composition of feature aggregation/propagation and transformation. Existing GCNs generally conduct feature aggregation on a fixed neighborhood graph in which each node computes its representation by aggregating the feature representations of all its neighbors (biased by its own representation). However, this fixed aggregation strategy is not guaranteed to be optimal for GCN based graph learning and also can be affected by some graph structure noises, such as incorrect or undesired edge connections. To address these issues, we propose a novel Graph mask Convolutional Network (GmCN) in which nodes can adaptively select the optimal neighbors in their feature aggregation to better serve GCN learning. More importantly, GmCN can be theoretically interpreted by a unified regularization framework, based on which we derive a simple update algorithm to determine the optimal mask adaptively in GmCN training process. Experiments on several datasets demonstrate the effectiveness of GmCN.
Graph Convolutional Networks (GCNs) have been widely studied for compact data representation and semi-supervised learning tasks. However, existing GCNs usually use a fixed neighborhood graph which is not guaranteed to be optimal for semi-supervised learning tasks. In this paper, we first re-interpret graph convolution operation in GCNs as a composition of feature propagation and (non-linear) transformation. Based on this observation, we then propose a unified adaptive neighborhood feature propagation model and derive a novel Adaptive Neighborhood Graph Propagation Network (ANGPN) for data representation and semi-supervised learning. The aim of ANGPN is to conduct both graph construction and graph convolution simultaneously and cooperatively in a unified formulation and thus can learn an optimal neighborhood graph that best serves graph convolution for data representation and semi-supervised learning. One main benefit of ANGPN is that the learned (convolutional) representation can provide useful weakly supervised information for constructing a better neighborhood graph which meanwhile facilitates data representation and learning. Experimental results on four benchmark datasets demonstrate the effectiveness and benefit of the proposed ANGPN.
RGB-Thermal object tracking attempt to locate target object using complementary visual and thermal infrared data. Existing RGB-T trackers fuse different modalities by robust feature representation learning or adaptive modal weighting. However, how to integrate dual attention mechanism for visual tracking is still a subject that has not been studied yet. In this paper, we propose two visual attention mechanisms for robust RGB-T object tracking. Specifically, the local attention is implemented by exploiting the common visual attention of RGB and thermal data to train deep classifiers. We also introduce the global attention, which is a multi-modal target-driven attention estimation network. It can provide global proposals for the classifier together with local proposals extracted from previous tracking result. Extensive experiments on two RGB-T benchmark datasets validated the effectiveness of our proposed algorithm.