@misc{indiciae22db5770552d,
  title = {Multi-Level LVLM Guidance for Untrimmed Video Action Recognition},
  author = {Liyang Peng and Sihan Zhu and Yunjie Guo},
  year = {2025},
  url = {https://arxiv.org/abs/2508.17442},
  note = {Source identifier: 2508.17442}
}