arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.00862cs.CVcs.AI
收藏

基于上下文和运动解耦的自监督视频表示学习

Self-supervised Video Representation Learning by Context and Motion Decoupling

Alibaba Group(阿里巴巴集团)

Lianghua Huang, Yu Liu, Bin Wang, Pan Pan, Yinghui Xu, Rong Jin

摘要:

自监督视频表示学习的一个关键挑战是如何有效地捕获除上下文偏差之外的运动信息。虽然大多数现有的作品隐含地通过视频特定的借口任务(例如,预测剪辑顺序、时间箭头和速度),我们开发了一种方法,该方法通过精心设计的借口任务明确地将运动监督从上下文偏差中分离出来。具体来说,我们将压缩视频中的关键帧和运动矢量(例如,H.264格式)分别作为上下文和运动的监督源,可以在CPU上以超过500 fps的速度有效地提取。然后,我们设计了两个借口的任务,联合优化:一个上下文匹配任务,成对的对比损失之间的视频剪辑和关键帧功能铸造;和一个运动预测任务,剪辑功能,通过编码器-解码器网络,用于估计运动功能在不久的将来。这两个任务使用共享的视频主干和单独的MLP头。实验表明,我们的方法提高了学习的视频表示的质量比以前的作品,其中我们获得了16.0%和11.1%的视频检索召回UCF 101和HMDB 51,分别绝对增益。此外,我们发现运动预测是视频网络的强正则化,其中使用它作为辅助任务提高了动作识别的准确性,幅度为7.4%~ 13.8%。

英文摘要:

A key challenge in self-supervised video representation learning is how to effectively capture motion information besides context bias. While most existing works implicitly achieve this with video-specific pretext tasks (e.g., predicting clip orders, time arrows, and paces), we develop a method that explicitly decouples motion supervision from context bias through a carefully designed pretext task. Specifically, we take the keyframes and motion vectors in compressed videos (e.g., in H.264 format) as the supervision sources for context and motion, respectively, which can be efficiently extracted at over 500 fps on the CPU. Then we design two pretext tasks that are jointly optimized: a context matching task where a pairwise contrastive loss is cast between video clip and keyframe features; and a motion prediction task where clip features, passed through an encoder-decoder network, are used to estimate motion features in a near future. These two tasks use a shared video backbone and separate MLP heads. Experiments show that our approach improves the quality of the learned video representation over previous works, where we obtain absolute gains of 16.0% and 11.1% in video retrieval recall on UCF101 and HMDB51, respectively. Moreover, we find the motion prediction to be a strong regularization for video networks, where using it as an auxiliary task improves the accuracy of action recognition with a margin of 7.4%~13.8%.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑