arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2302.03024cs.CV
收藏

AIM:自适应图像模型用于有效的视频动作识别

AIM: Adapting Image Models for Efficient Video Action Recognition

University of Central Florida(中佛罗里达大学);Amazon Web Services(亚马逊云科技)

Taojiannan Yang, Yi Zhu, Yusheng Xie, Aston Zhang, Chen Chen, Mu Li

摘要:

最近的基于视觉Transformer的视频模型大多遵循“图像预训练然后微调”的范式,并在多个视频基准测试中取得了巨大的成功。然而,考虑到预训练的图像Transformer模型已经表现出出色的可移植性,完全微调这样的视频模型在计算上可能是昂贵的并且是不必要的。在这项工作中,我们提出了一种新的方法来适应预训练的图像模型(AIM)有效的视频理解。通过冻结预训练的图像模型并添加一些轻量级的Adapters,我们引入了空间自适应,时间自适应和联合自适应,逐渐使图像模型具有时空推理能力。我们表明,我们提出的AIM可以实现竞争力,甚至比现有技术更好的性能与四个视频动作识别基准少得多的可调参数。由于其简单性,我们的方法通常也适用于不同的图像预训练模型,这有可能在未来利用更强大的图像基础模型。该项目的网页是\url{https://adapt-image-models.github.io/}。

英文摘要:

Recent vision transformer based video models mostly follow the ``image pre-training then finetuning" paradigm and have achieved great success on multiple video benchmarks. However, full finetuning such a video model could be computationally expensive and unnecessary, given the pre-trained image transformer models have demonstrated exceptional transferability. In this work, we propose a novel method to Adapt pre-trained Image Models (AIM) for efficient video understanding. By freezing the pre-trained image model and adding a few lightweight Adapters, we introduce spatial adaptation, temporal adaptation and joint adaptation to gradually equip an image model with spatiotemporal reasoning capability. We show that our proposed AIM can achieve competitive or even better performance than prior arts with substantially fewer tunable parameters on four video action recognition benchmarks. Thanks to its simplicity, our method is also generally applicable to different image pre-trained models, which has the potential to leverage more powerful image foundation models in the future. The project webpage is \url{https://adapt-image-models.github.io/}.

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。

↑