Deep subspace clustering uses latent features instead of raw images to construct the self-expression coefficient matrix. Existing methods primarily focus on optimizing the self-expression coefficient matrix, often neglecting the impact of latent features. However, better latent features are more in line with the self-representation assumption and results in a better self-expression coefficient matrix, which construct a chain relationship. Based on the chain relationship, this paper proposes a Class Relation Constraint (CRC) induced Deep Subspace Clustering (DSC) method to improve the representation ability of latent features. First, an intra- and inter-class weighted constraint is proposed to enhance latent data separability in subspaces. Then, to further remove negative samples inside a subspace, a contrastive loss function is introduced within the diagonal blocks of the self-expression coefficient matrix, i.e. the same subspace, under the guidance of spectral clustering results. Along with the enhanced representation ability on latent features and corresponding diagonal blocks, the self-expression coefficient matrix can provide more accurate data relationships for spectral clustering. Experimental results on multiple benchmark datasets have validated the effectiveness of the proposed DSCCRC method, particularly in handling small samples and complex datasets.

Unsupervised person re-identification (ReID), including fully unsupervised ReID and unsupervised domain adaptive ReID, remains a challenge for the fields of biometrics and computer vision due to its difficulty in learning with unlabeled target domain data. Existing state-of-the-art methods, most of which generate pseudo-labels via unsupervised clustering for model optimization, are inevitably hampered by the under-explored problem of pseudo-label noise. Motivated by this, we propose a novel joint framework termed pseudo-label Noise Prevention, Suppression, and Softening (NPSS) for unsupervised person re-identification. Instead of refining generated label noise after clustering as many existing methods do, we start solving this issue from the source of pseudo-label noise by proposing a new Dynamic Camera-Adaptive Clustering (DCAC), which dynamically involves camera information to prevent noise caused by cross-camera variance, thus improving their quality during clustering. Moreover, we propose an Online Domain Union (ODU) mechanism for the classification model learning on the target domain via involving source domain data with their ground-truth labels, which effectively suppresses the indelible noisy pseudo-labels. Furthermore, we present the Self-Consistency Constraint (SCC) to soften the label noise in a single model with reduced computation and network parameter cost, which achieves intra-sample knowledge ensembling with our global-local SCC and cross-sample knowledge ensembling with our inter-instance SCC. Experiments demonstrate the effectiveness of our method as it surpasses state-of-the-art methods by a large margin on Market-1501, DukeMTMC-ReID, and MSMT17 benchmarks. The code is available at https://github.com/hjwang-824/NPSS.

Domain generalizable (DG) semantic segmentation, i.e., a semantic segmentation model pretrained from a source domain performs well in previously unseen target domains without any fine-tuning, remains an open question. A promising solution is learning style-agnostic and domain-invariant features with stylized augmented data. However, existing methods mainly focused on performing stylization on coarse-grained image-level features, while ignoring to explore fine-grained semantic style clues and high-order semantic context correlation, which are essential in enhancing the generalization. Motivated by this, we propose a novel framework termed Consistent Learning based on Class-Aware Style Variation (CL-CASV) for DG semantic segmentation. Specifically, with the guidance of class-level semantic information, our proposed Class-Aware Style Variation (CASV) module simulates imaging object and imaging condition style variation that can appear in complex real-world scenarios, thus generating fine-grained class-aware stylized images with rich style variation. Then the similarities between augmentations and original images are exploited via our Self-Correlation Consistency Learning (SCCL) that mines global context consistency from the views of channel correlation and spatial correlation in the feature and prediction spaces. Extensive experiments on mainstream benchmarks, including Cityscapes, GTAV, BDD100K, SYNTHIA, and Mapillary, demonstrate the effectiveness of our method as it surpasses the state-of-the-art methods.

Synthetic aperture radar (SAR) image segmentation is a challenge due to its inherent speckle. Gamma distribution is believed to be an appropriate statistical model to describe the characteristics of speckle in SAR images. In this letter, a fuzzy clustering algorithm based on gamma distribution for SAR image segmentation is proposed, in which the Stirling equation is used to approach the gamma function in the dominator of gamma distribution under the assumption of mean field theory to make the shape parameter$\alpha $derivable. Then, the value range of the estimated$\alpha $is demonstrated to meet the requirement of gamma distribution by Jensen’s inequality. Experimental results show that the proposed method gives promising results in SAR image (including large area SAR image) segmentation and effectively suppresses the influence of speckle.