USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation
Proceedings of the AAAI Conference on Artificial Intelligence · 11.04.2025 · 39 (8) · 8332-8340
DOI: 10.1609/aaai.v39i8.32899 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 11.04.2025. Verfasst von Wanjiang Weng und 4 weiteren Autor:innen. Die Arbeit wurde bisher 18-mal zitiert (Crossref).
Abstract (Originalsprache)
Contrastive learning has achieved great success in skeleton-based representation learning recently. However, the prevailing methods are predominantly negative-based, necessitating additional momentum encoder and memory bank to get negative samples, which increases the difficulty of model training. Furthermore, these methods primarily concentrate on learning a global representation for recognition and retrieval tasks, while overlooking the rich and detailed local representations that are crucial for dense prediction tasks. To alleviate these issues, we introduce a Unified Skeleton-based Dense Representation Learning framework based on feature decorrelation, called USDRL, which employs feature decorrelation across temporal, spatial, and instance domains in a multi-grained manner to reduce redundancy among dimensions of the representations to maximize information extraction from features. Additionally, we design a Dense Spatio-Temporal Encoder (DSTE) to capture fine-grained action representations effectively, thereby enhancing the performance of dense prediction tasks. Comprehensive experiments, conducted on the benchmarks NTU-60, NTU-120, PKU-MMD I, and PKU-MMD II, across diverse downstream tasks including action recognition, action retrieval, and action detection, conclusively demonstrate that our approach significantly outperforms the current state-of-the-art (SOTA) approaches.
Zitieren
APA
Weng, W., Wang, H., Wang, J., He, L., & Xie, G. (2025). USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation. Proceedings of the AAAI Conference on Artificial Intelligence, 39(8), 8332-8340. https://doi.org/10.1609/aaai.v39i8.32899
Vancouver
Weng W, Wang H, Wang J, He L, Xie G. USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation. Proceedings of the AAAI Conference on Artificial Intelligence. 2025;39(8):8332-8340. doi:10.1609/aaai.v39i8.32899
BibTeX
@article{Weng2025,
author = {Weng, Wanjiang and Wang, Hongsong and Wang, Junbo and He, Lei and Xie, Guo-Sen},
title = {USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2025},
volume = {39},
number = {8},
pages = {8332--8340},
doi = {10.1609/aaai.v39i8.32899}
}
EUVIMED