Permutation-Aware Action Segmentation via Unsupervised Frame-to-Segment Alignment

TL;DR

This paper introduces an unsupervised transformer-based framework for activity segmentation that combines frame-level and segment-level cues, utilizing temporal optimal transport and frame-to-segment alignment to improve segmentation accuracy.

Contribution

It proposes a novel permutation-aware, unsupervised approach that integrates frame-to-segment alignment with transformer models for activity segmentation.

Findings

Achieves comparable or better performance than previous methods on four public datasets.

Utilizes temporal optimal transport for unsupervised training.

Introduces pseudo labels for effective unsupervised learning.

Abstract

This paper presents an unsupervised transformer-based framework for temporal activity segmentation which leverages not only frame-level cues but also segment-level cues. This is in contrast with previous methods which often rely on frame-level information only. Our approach begins with a frame-level prediction module which estimates framewise action classes via a transformer encoder. The frame-level prediction module is trained in an unsupervised manner via temporal optimal transport. To exploit segment-level information, we utilize a segment-level prediction module and a frame-to-segment alignment module. The former includes a transformer decoder for estimating video transcripts, while the latter matches frame-level features with segment-level features, yielding permutation-aware segmentation results. Moreover, inspired by temporal optimal transport, we introduce simple-yet-effective pseudo labels for unsupervised training of the above modules. Our experiments on four public datasets, i.e., 50 Salads, YouTube Instructions, Breakfast, and Desktop Assembly show that our approach achieves comparable or better performance than previous methods in unsupervised activity segmentation. Our code and dataset are available on our research website: https://retrocausal.ai/research/.