EUVIMEDEUVIMED entdecken
Crossref

CSTA: Spatial-Temporal Causal Adaptive Learning for Exemplar-Free Video Class-Incremental Learning

Tieyuan Chen, Huabin Liu, Chern Hong Lim, John See, Xing Gao, Junhui Hou, Weiyao Lin

IEEE Transactions on Circuits and Systems for Video Technology · 2025-11 · 35 (11) · 11488-11501

DOI: 10.1109/tcsvt.2025.3579477 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2025-11. Verfasst von Tieyuan Chen und 6 weiteren Autor:innen. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 63 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Shanghai “The Belt and Road” Young Scholar Exchange, Zhongguancun Academy Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Continual learning aims to acquire new knowledge while retaining past information. Class-incremental learning (CIL) presents a challenging scenario where classes are introduced sequentially. For video data, the task becomes more complex than image data because it requires learning and preserving both spatial appearance and temporal action involvement. To address this challenge, we propose a novel exemplar-free framework that equips separate spatiotemporal adapters to learn new class patterns, accommodating the incremental information representation requirements unique to each class. While separate adapters are proven to mitigate forgetting and fit unique requirements, naively applying them hinders the intrinsic connection between spatial and temporal information increments, affecting the efficiency of representing newly learned class information. Motivated by this, we introduce two key innovations from a causal perspective. First, a causal distillation module is devised to maintain the relation between spatial-temporal knowledge for a more efficient representation. Second, a causal compensation mechanism is proposed to reduce the conflicts during increment and memorization between different types of information. Extensive experiments conducted on benchmark datasets demonstrate that our framework can achieve new state-of-the-art results, surpassing current example-based methods by 4.2% in accuracy on average. The codes are accessible in https:// github.com/tychen-SJTU/CSTA.

Förderung

  • National Natural Science Foundation of China
  • Shanghai “The Belt and Road” Young Scholar Exchange
  • Zhongguancun Academy Project
  • Malaysia Ministry of Higher Education

Zitieren

APA

Chen, T., Liu, H., Lim, C. H., See, J., Gao, X., Hou, J., & Lin, W. (2025). CSTA: Spatial-Temporal Causal Adaptive Learning for Exemplar-Free Video Class-Incremental Learning. IEEE Transactions on Circuits and Systems for Video Technology, 35(11), 11488-11501. https://doi.org/10.1109/tcsvt.2025.3579477

Vancouver

Chen T, Liu H, Lim CH, See J, Gao X, Hou J, et al. CSTA: Spatial-Temporal Causal Adaptive Learning for Exemplar-Free Video Class-Incremental Learning. IEEE Transactions on Circuits and Systems for Video Technology. 2025;35(11):11488-11501. doi:10.1109/tcsvt.2025.3579477

BibTeX
@article{Chen2025,
  author = {Chen, Tieyuan and Liu, Huabin and Lim, Chern Hong and See, John and Gao, Xing and Hou, Junhui and Lin, Weiyao},
  title = {CSTA: Spatial-Temporal Causal Adaptive Learning for Exemplar-Free Video Class-Incremental Learning},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  year = {2025},
  volume = {35},
  number = {11},
  pages = {11488--11501},
  doi = {10.1109/tcsvt.2025.3579477}
}
Kostenlos starten →