Understanding Video Transformers via Universal Concept Discovery

Matthew Kowal, Achal Dave, Rares Andrei Ambrus, Adrien Gaidon, Konstantinos G. Derpanis, Pavel Tokmakov

Keywords: Interpretability, Video Understanding, Object-Centric Representations, Universal Concepts

TL;DR: We discover universal concepts in video transformers that provide understanding of general representations learned across tasks.

Abstract:

This paper studies the problem of concept-based interpretability of transformer representations for videos. Concretely, we seek to explain the decision-making process of video transformers based on high-level, spatiotemporal concepts that are automatically discovered.
Prior research on concept-based interpretability has concentrated solely on image-level tasks. Comparatively, video models deal with the added temporal dimension, increasing complexity and posing challenges in identifying dynamic concepts over time. In this work, we systematically address these challenges by introducing the first Video Transformer Concept Discovery (VTCD) algorithm. To this end, we propose an efficient approach for unsupervised identification of units of video transformer representations - concepts, and ranking their importance to the output of a model. The resulting concepts are highly interpretable, revealing spatio-temporal reasoning mechanisms and object-centric representations in unstructured video models. Performing this analysis jointly over a diverse set of supervised and self-supervised representations, we discover that some of these mechanism are universal in video transformers. Finally, we show that VTCD can be used for fine-grained action recognition and video object segmentation.

Submission Number: 14