Language model guided interpretable video action reasoning
Abstract
While neural networks have excelled in video action recognition tasks, their 'black-box' nature often obscures the understanding of their decision-making processes. Re-cent approaches used inherently interpretable models to an-alyze video actions in a manner akin to human reasoning. These models, however, usually fall short in performance compared to their 'black-box' counterparts. In this work, we present a new framework named Language-guided Interpretable Action Recognition framework (La-IAR). LaIAR leverages knowledge from language models to enhance both the recognition capabilities and the inter-pretability of video models. In essence, we redefine the problem of understanding video model decisions as a task of aligning video and language models. Using the logical reasoning captured by the language model, we steer the training of the video model. This integrated approach not only improves the video model's adaptability to different domains but also boosts its overall performance. Extensive experiments on two complex video action datasets, Charades & CAD-120, validates the improved performance and inter-pretability of our LaIAR framework. The code of LaIAR is available at https://github.com/NingWang2049/LaIAR.
RAS ID
65876
Document Type
Conference Proceeding
Date of Publication
1-1-2024
Funding Information
National Natural Science Foundation of China (62073252, 62072358) / Natural Science Basic Research Program of Shaanxi (2024JC-JCQN-66)
School
School of Science
Copyright
subscription content
Publisher
IEEE
Identifier
Syed Afaq Ali Shah: https://orcid.org/0000-0003-2181-8445
Comments
Wang, N., Zhu, G., Li, H., Zhang, L., Ali Shah, S. A., & Bennamoun, M. (2024). Language model guided interpretable video action reasoning. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 18878-18887). IEEE. https://doi.org/10.1109/CVPR52733.2024.01786