Haoyuan Zhang

Haoyuan Zhang


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications

28 February 2025 in Sensors
Verified
Journal Article
Open access
PDF file
HZHaoyuan Zhang
Haoyuan Zhang
In this paper, we propose a contrastive mask learning (CML) method for self-supervised 3D skeleton-based action recognition. Specifically, the mask modeling mechanism is integrated into multi-level contrastive learning with the aim of forming a mutually beneficial learning scheme from both contrastive learning and masked skeleton reconstruction. The contrastive objective is extended from an individual skeleton instance to clusters by closing the gap between cluster assignment from different instances of the same category, with the goal of pursuing inter-instance consistency. Compared with previous methods, CML integrates contrastive and masked learning comprehensively and enables intra-/inter-instance consistency pursuit via multi-level contrast, which leads to more discriminative skeleton representation learning. Our extensive evaluation of the challenging NTU RGB+D and PKU-MMD benchmarks demonstrates that representations learned via CML exhibit superior discriminability, consistently outperforming state-of-the-art methods in terms of action recognition accuracy.

19 July 2024 in Complex & Intelligent Systems
Verified
Journal Article
Open access
PDF file
HZHaoyuan Zhang
Haoyuan Zhang
Most existing 3D action recognition works rely on the supervised learning paradigm, yet the limited availability of annotated data limits the full potential of encoding networks. As a result, effective self-supervised pre-training strategies have been actively researched. In this paper, we target to explore a self-supervised learning approach for 3D action recognition, and propose the Attention-guided Mask Learning (AML) scheme. Specifically, the dropping mechanism is introduced into contrastive learning to develop Attention-guided Mask (AM) module as well as mask learning strategy, respectively. The AM module leverages the spatial and temporal attention to guide the corresponding features masking, so as to produce the masked contrastive object. The mask learning strategy enables the model to discriminate different actions even with important features masked, which makes action representation learning more discriminative. What’s more, to alleviate the strict positive constraint that would hinder representation learning, the positive-enhanced learning strategy is leveraged in the second-stage training. Extensive experiments on NTU-60, NTU-120, and PKU-MMD datasets show that the proposed AML scheme improves the performance in self-supervised 3D action recognition, achieving state-of-the-art results.

29 October 2020 in arXiv
Verified
Preprint
Open access
PDF file
HZHaoyuan Zhang
Haoyuan Zhang
YHYonghong Hou
Yonghong Hou
PWPichao WANG
Pichao WANG
ZGZihui Guo
Zihui Guo
WLWanqing Li
Wanqing Li
This paper presents a study of automatic design of neural network architectures for skeleton-based action recognition. Specifically, we encode a skeleton-based action instance into a tensor and carefully define a set of operations to build two types of network cells: normal cells and reduction cells. The recently developed DARTS (Differentiable Architecture Search) is adopted to search for an effective network architecture that is built upon the two types of cells. All operations are 2D based in order to reduce the overall computation and search space. Experiments on the challenging NTU RGB+D and Kinectics datasets have verified that most of the networks developed to date for skeleton-based action recognition are likely not compact and efficient. The proposed method provides an approach to search for such a compact network that is able to achieve comparative or even better performance than the state-of-the-art methods.