@misc{indiciaec399813eacbd,
  title = {MoCrop: Training Free Motion Guided Cropping for Efficient Video Action Recognition},
  author = {Binhua Huang and Wendong Yao and Shaowu Chen and Guoxin Wang and Qingyuan Wang and Soumyabrata Dev},
  year = {2026},
  url = {https://arxiv.org/abs/2509.18473},
  note = {Source identifier: 2509.18473}
}