In this report, we propose a video-language pretraining (VLP) based solution \cite{kevin2022egovlp} for the EPIC-KITCHENS-100 Multi-Instance Retrieval (MIR) challenge. Especially, we exploit the recently released Ego4D dataset \cite{grauman2021ego4d} to pioneer Egocentric VLP from pretraining dataset, pretraining objective, and development set. Based on the above three designs, we develop a pretrained video-language model that is able to transfer its egocentric video-text representation to MIR benchmark. Furthermore, we devise an adaptive multi-instance max-margin loss to effectively fine-tune the model and equip the dual-softmax technique for reliable inference. Our best single model obtains strong performance on the challenge test set with 47.39% mAP and 61.44% nDCG. The code is available at https://github.com/showlab/EgoVLP.
@article{arxiv.2207.01334,
title = {Egocentric Video-Language Pretraining @ EPIC-KITCHENS-100 Multi-Instance Retrieval Challenge 2022},
author = {Kevin Qinghong Lin and Alex Jinpeng Wang and Rui Yan and Eric Zhongcong Xu and Rongcheng Tu and Yanru Zhu and Wenzhe Zhao and Weijie Kong and Chengfei Cai and Hongfa Wang and Wei Liu and Mike Zheng Shou},
journal= {arXiv preprint arXiv:2207.01334},
year = {2022}
}
Comments
To appeared in CVPRW22. 5 pages, 2 figures, 2 tables. Code: https://github.com/showlab/EgoVLP. The EPIC challenge technical report of EgoVLP arXiv:2206.01670. See Ego4D challenge technical report arXiv:2207.01622