Transformers in Action: Weakly Supervised Action Segmentation

TL;DR

This paper explores the application of transformers to weakly supervised video action segmentation, demonstrating improved accuracy and efficiency over RNN-based models through novel architecture and transcript embedding techniques.

Contribution

It introduces a transformer-based architecture tailored for weakly supervised action segmentation and a transcript embedding method for faster inference and better segmentation performance.

Findings

Transformers outperform RNNs in action alignment accuracy.

The proposed transcript embedding accelerates inference.

The approach improves segmentation results on benchmark datasets.

Abstract

The video action segmentation task is regularly explored under weaker forms of supervision, such as transcript supervision, where a list of actions is easier to obtain than dense frame-wise labels. In this formulation, the task presents various challenges for sequence modeling approaches due to the emphasis on action transition points, long sequence lengths, and frame contextualization, making the task well-posed for transformers. Given developments enabling transformers to scale linearly, we demonstrate through our architecture how they can be applied to improve action alignment accuracy over the equivalent RNN-based models with the attention mechanism focusing around salient action transition regions. Additionally, given the recent focus on inference-time transcript selection, we propose a supplemental transcript embedding approach to select transcripts more quickly at inference-time. Furthermore, we subsequently demonstrate how this approach can also improve the overall segmentation performance. Finally, we evaluate our proposed methods across the benchmark datasets to better understand the applicability of transformers and the importance of transcript selection on this video-driven weakly-supervised task.