/HME-VideoQA

Heterogeneous Memory Enhanced Multimodal Attention Model for VideoQA

Primary LanguagePython

Heterogeneous Memory Enhanced Multimodal Attention Model for VideoQA

(HME-VideoQA)

This is the PyTorch Implementation of

  • Chenyou Fan, Xiaofan Zhang, Shu Zhang, Wensheng Wang, Chi Zhang, Heng Huang. Heterogeneous Memory Enhanced Multimodal Attention Model for VideoQA. In CVPR, 2019. [link]
@inproceedings{fan-CVPR-2019,
   author    = {Chenyou Fan, Xiaofan Zhang, Shu Zhang, Wensheng Wang, Chi Zhang, Heng Huang},
   title     = "{Heterogeneous Memory Enhanced Multimodal Attention Model for Video Question Answering}"
   booktitle = {CVPR},
   year      = 2019
}

VideoQA Task

Task

Architecture

Network

Datasets

TGIF-QA, see gif-qa/

MSVD-QA, see msvd-qa/

Youtube2text, see zh-qa/

Requirements

Python = 2.7

PyTorch = 1.0 [here]

GPU training with 4G+ memory, testing with 1G+ memory.