Exploring Explainability in Video Action Recognition

TL;DR

This paper investigates explainability methods for video action recognition, introduces Video-TCAV for concept importance quantification, and proposes a machine-assisted approach for generating relevant spatiotemporal concepts.

Contribution

It extends TCAV to video tasks, addresses concept generation challenges, and emphasizes the importance of dynamic spatiotemporal concepts in explainability.

Findings

Dynamic spatiotemporal concepts outperform static spatial ones.

Video-TCAV effectively quantifies concept importance in video models.

Proposed machine-assisted concept generation aids explainability research.

Abstract

Image Classification and Video Action Recognition are perhaps the two most foundational tasks in computer vision. Consequently, explaining the inner workings of trained deep neural networks is of prime importance. While numerous efforts focus on explaining the decisions of trained deep neural networks in image classification, exploration in the domain of its temporal version, video action recognition, has been scant. In this work, we take a deeper look at this problem. We begin by revisiting Grad-CAM, one of the popular feature attribution methods for Image Classification, and its extension to Video Action Recognition tasks and examine the method's limitations. To address these, we introduce Video-TCAV, by building on TCAV for Image Classification tasks, which aims to quantify the importance of specific concepts in the decision-making process of Video Action Recognition models. As the scalable generation of concepts is still an open problem, we propose a machine-assisted approach to generate spatial and spatiotemporal concepts relevant to Video Action Recognition for testing Video-TCAV. We then establish the importance of temporally-varying concepts by demonstrating the superiority of dynamic spatiotemporal concepts over trivial spatial concepts. In conclusion, we introduce a framework for investigating hypotheses in action recognition and quantitatively testing them, thus advancing research in the explainability of deep neural networks used in video action recognition.