@misc{indiciae998ffd393330,
  title = {Denoise and Align: Diffusion-Driven Foreground Knowledge Prompting for Open-Vocabulary Temporal Action Detection},
  author = {Sa Zhu and Wanqian Zhang and Lin Wang and Jinchao Zhang and Cong Wang and Bo Li},
  year = {2026},
  url = {https://arxiv.org/abs/2604.18313},
  note = {Source identifier: 2604.18313}
}