AAAI 2026technical0 citations

Learning Dynamics as Feedback: An Adaptive Entropy Flow Dynamics Framework for Long-tailed Human Action Recognition

Yuan Dong, Zhe Zhao, Liheng Yu, Di Wu, Pengkun Wang

Abstract

Deep human action recognition models trained on real-world data are often challenged by long-tailed distributions, where performance on rare classes is severely degraded. Current solutions typically apply static or heuristic interventions that are disconnected from the model

BibTeX
@inproceedings{aaai2026_learningdynamics,
  title = {Learning Dynamics as Feedback: An Adaptive Entropy Flow Dynamics Framework for Long-tailed Human Action Recognition},
  author = {Yuan Dong and Zhe Zhao and Liheng Yu and Di Wu and Pengkun Wang},
  booktitle = {AAAI 2026},
  year = {2026}
}