Feature Selection

Joint image clustering and feature selection with auto-adjoined learning for high-dimensional data, Knowledge-Based Systems, and code is here
Xiaodong Wang, Pengtao Wu, Qinghua Xu,Zhiqiang Zeng, and Yong Xie;

Abstract: We impose an extremely sparse feature selection matrix into K-means, which is easy to be optimized. Besides, to accurately encode the local adjacency among data without the influence of noise, we propose to automatically assign the connectivity of each sample in the low-dimensional feature space.

Fast Adaptive K-Means Subspace Clustering for High-Dimensional Data, IEEE ACCESS, and code is here
Xiaodong Wang, Rungching Chen, Fei Yan,Zhiqiang Zeng, and Chaoqun Hong;
Adaptive multi-view subspace clustering for high-dimensional data, Pattern Recognition Letters, and code is here
Fei Yan, Xiaodong Wang, Zhiqiang Zeng, Chaoqun Hong

Abstract: We propose a new adaptive multi-view subspace clustering method to integrate heterogenous data in the low-dimensional feature space.

Semi-supervised multi-label feature selection via label correlation analysis with l1-norm graph embedding,Image and Vision Computing, and code is here.
Xiaodong Wang, Rungching Chen, Chaoqun Hong, Zhiqiang Zeng, Zhili Zhou;

Abstract: We propose a novel semi-supervised multi-label feature selection algorithm and apply it to three different applications: natural scene classification, web page annotation, and yeast gene functional classification.

Robust Dimension Reduction for Clustering With Local Adaptive Learning, Pattern Recognition Letters.
Xiaodong Wang, Rungching Chen, Zhiqiang Zeng, Chaoqun Hong, Fei Yan;

Abstract: We explore the discriminative information among data by jointly unifying local adaptive subspace learning and KM clustering.

Semi-supervised adaptive feature analysis and its application for multimedia understanding, Multimedia Tools and Applications.
Xiaodong Wang, Rungching Chen, Fei Yan, Zhiqiang Zeng, Chaoqun Hong;

Abstract: In this paper, adaptive local manifold learning and feature selection are integrated jointly into a single framework.

Local adaptive learning for semi-supervised feature selection with group sparsity, Knowledge-Based Systems, and code is here
Zhiqiang Zeng, Xiaodong Wang, Fei Yan, Yuming Chen;

Abstract: The proposed method explores the local sparsity of data through a local and global regression learning manner.

Semi-supervised feature selection with exploiting shared information among multiple tasks, J. Vis. Commun. Image R.
Xiaodong Wang, Rungching Chen, Fei Yan, Zhiqiang Zeng;

Abstract: we propose a novel feature selection algorithm, which integrates the semi-supervised learning and multi-task learning into a joint framework.

Unsupervised spectral feature selection with l1-norm graph, Neurocomputing.
Xiaodong Wang, Xu Zhang, Zhiqiang Zeng, Qun Wu,Jian Zhang;

Abstract: We propose a new Unsupervised Spectral Feature Selection with l1-Norm Graph, namely USFS. It performs the spectral clustering and l1-Norm Graph jointly to select discriminative features.

Network Pruning

Progressive Local Filter Pruning for Image Retrieval Acceleration, IEEE Transactions on Multimedia.
Xiaodong Wang, Zhedong Zheng, Yang He, Fei Yan, Zhiqiang Zeng, Yi Yang;

Abstract: We propose a local progressive filter pruning method for image retrieval networks.

Soft person reidentification network pruning via blockwise adjacent filter decaying, IEEE Transactions on Cybernetics, and code is here
Xiaodong Wang, Zhedong Zheng, Yang He, Fei Yan, Zhiqiang Zeng, Yi Yang;

Abstract: We propose a blockwise adjacent filter decaying method to accelerate deep re-id networks that produce continuous features and are sensitive to network pruning.