In this report, we present our champion solutions to five tracks at Ego4D challenge. We leverage our developed InternVideo, a video foundation model, for five Ego4D tasks, including Moment Queries, Natural Language Queries, Future Hand Prediction, State Change Object Detection, and Short-term Object Interaction Anticipation. InternVideo-Ego4D is an effective paradigm to adapt the strong foundation model to the downstream ego-centric video understanding tasks with simple head designs. In these five tasks, the performance of InternVideo-Ego4D comprehensively surpasses the baseline methods and the champions of CVPR2022, demonstrating the powerful representation ability of InternVideo as a video foundation model. Our code will be released at https://github.com/OpenGVLab/ego4d-eccv2022-solutions
@article{arxiv.2211.09529,
title = {InternVideo-Ego4D: A Pack of Champion Solutions to Ego4D Challenges},
author = {Guo Chen and Sen Xing and Zhe Chen and Yi Wang and Kunchang Li and Yizhuo Li and Yi Liu and Jiahao Wang and Yin-Dong Zheng and Bingkun Huang and Zhiyu Zhao and Junting Pan and Yifei Huang and Zun Wang and Jiashuo Yu and Yinan He and Hongjie Zhang and Tong Lu and Yali Wang and Limin Wang and Yu Qiao},
journal= {arXiv preprint arXiv:2211.09529},
year = {2022}
}
Comments
Technical report in 2nd International Ego4D Workshop@ECCV 2022. Code will be released at https://github.com/OpenGVLab/ego4d-eccv2022-solutions