This is the code for our solution in first ICMI-MAVSR challenge track 1 and track 3. If you find this code useful for your research, please kindly cite:
@inproceedings{yao2019spotting,
title={Spotting Visual Keywords from Temporal Sliding Windows},
author={Yao, Yue and Wang, Tianyu and Du, Heming and Zheng, Liang and Gedeon, Tom},
booktitle={2019 International Conference on Multimodal Interaction},
pages={536--539},
year={2019},
organization={ACM}
}