Multi-view clustering aims to partition multi-view data into different categories by optimally exploring the consistency and complementary information from multiple sources. However, most existing multi-view clustering algorithms heavily rely on the similarity graphs from respective views and fail to comprehend multiple views holistically. Moreover, due to the noise and redundancy maintained in the original data, the original errors of multiple similarity graphs will continue to accumulate in the process of constructing consistent graphs. These situations always lead to the limitation to effective fuse the essential information from multiple views, which always influences the clustering performance and cries out for reliable solutions. Based on the above considerations, we propose a novel method termed Tensorial Multi-view Clustering (TMvC), which learns high-order graph by low-rank tensor constraint to uncover the essential information stored in multiple views. TMvC first learns the Laplacian graphs of all views and stacks them into a tensor which can be viewed as a high-order graph. With the high-order graph, consistency and complementary information from different views can be propagated smoothly across all views. Then, based on low-rank constraint, high-order graph is constrained in the horizontal and vertical directions to better uncover the inter-view and inter-class correlations between multi-view data, which is of vital importance for multi-view clustering. Extensive experiments on document and image datasets demonstrate that TMvC can achieve the state-of-the-art performance for multi-view clustering.

Vehicle re-identification (reID) targets at matching same vehicle images captured from multi-cameras, which has become a hot topic in recent years. However, it poses the critical but challenging problem that is caused by complicated environments and diversified illuminations. Inspired by the benefit of attributes that contain detailed descriptions, this paper proposes a novel Attribute-Guided Network (AGNet), which learns global representation with the abundant attribute features in an end-to-end manner. Specially, an attribute-guided module is proposed in AGNet to generate the attribute mask which inversely guide to select discriminative features for category classification. Moreover, in AGNet, an attribute-based label smoothing loss is presented to better train the reID model, which can strength the distinct ability of reID model and regularize AGNet model according to the attributes. The experiment result validate the effectiveness of AGNet and demonstrate AGNet achieves excellent performance on two benchmark datasets.

Re-identification (re-ID) aims to identify a person’s images across different cameras. However, the domain differences between different datasets make it a challenge for re-ID models trained on one dataset to be adapted to another. A variety of unsupervised domain adaptation methods tend to transfer learned knowledge from one domain to another by optimizing with pseudo-labels. Though impressive performances have been achieved, there are still some limitations. To be specific, these methods always generate one pseudo label for each unlabeled sample, which is hard to describe a person accurately and introduces a large number of noisy labels by one-shot clustering, thus hindering the retraining process and limiting generalization. To build more comprehensive descriptions of samples and mitigate the effects of noisy pseudo labels, this paper proposes an Adaptive Memorization with Group labels (AdaMG) framework for unsupervised person re-ID, which resists noisy labels and exploits the diversity of samples by developing a multi-branch structure with the adaptive memorization. In particular, group labels are generated for one sample in the unseen domain to learn more complementary and diverse features through clustering. Meanwhile, to better optimize the neural networks with noisy data, multiple memory structures are designed in AdaMG, which are updated adaptively according to the confidence of samples. Comprehensive experimental results have demonstrated that our proposed method can achieve excellent performances on benchmark datasets.

The Multi-view clustering aims to cluster multi-view data into distinct clusters by exploring consistency and complementary information from different sources. However, most of the existing spectral clustering methods divide samples by similarity relationships, ignoring the higher-order information relationships of the underlying clusters in different views. Meanwhile, the performance of these methods involves the control of multiple hyperparameters, as multiple hyperparameters interact with each other, that tends to lead to uncontrollable performance of the methods, which further limits the practical application of the models. Existing methods treat the sample information of each view equally in the sample clustering process, which is not reasonable in real-world applications. The important information contained in each view is different, and treating each view equally will reduce the contribution of important information and increase the contribution of minor information, leading to a decrease in clustering performance. To solve these problems, this paper proposes a new spectral clustering model called Tensorized Parameter-free Multi-view Spectral Clustering based on Fair Representation Learning(TPSC). First, TPSC obtains spectral embedding matrices from different views and stacks them into tensors to mine the clustering relationships of the samples from different perspectives, as well as to explore the higher-order consistency information of each spectral embedding matrix. In addition, TPSC provides adaptive weights for the samples of each view to improve the robustness of the algorithm. The consistency relationships of different views can be mined using this strategy, and the intra-cluster relationships of sample points can be balanced from various perspectives. Finally, TPSC designs an efficient optimisation algorithm to solve the resulting optimisation problem. Extensive experiments are conducted on several baseline datasets to confirm the effectiveness of the approach.