Kang Wei

Mr. Kang Wei

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 May 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
XGXizhan Gao
Xizhan Gao
KWKang Wei
Kang Wei
JLJia Li
Jia Li
ZSZiyu Shi
Ziyu Shi
HZHui Zhao
Hui Zhao
SNSijie Niu
Sijie Niu
As an important research direction in image and video processing, set-based video recognition requires speed and accuracy. However, the existing static modeling methods focus on computational speed but ignore accuracy, whereas the dynamic modeling methods are higher-accuracy but ignore the computational speed. Combining these two types of methods to obtain fast and accurate recognition results remains a challenging problem. Motivated by this, in this study, a novel Manifolds-based Low-Rank Dictionary Pair Learning (MbLRDPL) method was developed for a set-based video recognition/image set classification task. Specifically, each video or image set was first modeled as a covariance matrix or linear subspace, which can be seen as a point on a Riemannian manifold. Second, the proposed MbLRDPL learned discriminative class-specific synthesis and analysis dictionaries by clearly imposing the nuclear norm on the synthesis dictionaries. The experimental results show that our method achieved the best classification accuracy (100%, 72.16%, 95%) on three datasets with the fastest computing time, reducing the errors of state-of-the-art methods (JMLC, DML, CEBSR) by 0.96–75.69%.

17 May 2024 in IEEE Signal Processing Letters
Verified
Journal Article
Closed access
XGXizhan Gao
Xizhan Gao
KWKang Wei
Kang Wei
SNSijie Niu
Sijie Niu
HZHui Zhao
Hui Zhao
JDJiwen Dong
Jiwen Dong
SAR image classification is one of the research hotspots in the field of remote sensing. The performance of SAR image classification greatly depends on the learning of features and the design of classifiers. However, traditional SAR image classification methods either focus on learning (deep) features or focus on designing discriminative classifiers, while ignoring the correlation between them, resulting in the learned features and classifiers often not matching. To address the above issues, in this paper a novel classifier, termed entropy based low-rank dictionary pair learning (ELRDPL) method is first designed, which introduces the entropy theory and low-rank constraints into the objective function of dictionary learning for increasing the discriminative ability and decreasing the space occupation. Based on the constructed classifier, an entropy based deep lowrank dictionary pair learning network (EDLRDPL Net) is then proposed, which performs joint learning of deep features and discriminative dictionaries by embedding the ELRDPL classifier into the deep neural network. Extensive experimental results on four SAR image classification datasets demonstrate the effectiveness of EDLRDPL Net.

24 February 2025 in IEEE Transactions on Biometrics, Behavior, and Identity Science
Verified
Journal Article
Closed access
XGXizhan Gao
Xizhan Gao
XLXingrui Liu
Xingrui Liu
WHWei Hu
Wei Hu
SNSijie Niu
Sijie Niu
KWKang Wei
Kang Wei
Danilo PelusiDanilo Pelusi
Danilo Pelusi
As an important biometric technology, video face identification has broad application prospects and has been widely concerned by researchers for a long time. Although many video face identification methods have emerged, they still have not fully overcome the inherent problems such as intra-video diversity, inter-video similarity. In view of this, a set-based video face identification method called Multi-level Feature Perception Network (MFPN) is proposed, which consists of two modules: a video feature generation module and a distance metric learning module. These modules fully consider the cohesiveness within each video and the similarity between different videos. At first, the video feature generation module is used to learn (or perceive) the atom-level features of each frame, and in this process, the cohesiveness problem within each video is considered to facilitate the subsequent learning of compact video features. Then, the distance metric learning module is utilized to further learn (or perceive) the deep concept-level features of videos taking into account the similarity between different videos to enhance the discriminative ability of these features. Finally, the class of the test video is determined based on its deep concept-level features. Experimental results on the Honda, MoBo, YTC, and YTF video face datasets show that the proposed method significantly improves recognition rates in both video face identification and video face verification tasks.