LoCo-MAD: Long-Range Context-Enhanced Model Towards Plot-Centric Movie Audio Description

Details

Results

Datasets

MAD-v2: please refer to AutoAD. The size of propressed dataset is over 50G. Please e-mail me to obtain it.

Reference

@InProceedings{wang2024loco-mad,
  title={LoCo-MAD: Long-Range Context-Enhanced Model Towards Plot-Centric Movie Audio Description},  
  author={Jiayi Wang and Zihao Liu and Xiaoyu Wu},  
  booktitle={ACCV},  
  year={2024}}