Do we really need temporal convolutions in action segmentation?

TL;DR

This paper introduces a pure Transformer-based model called Temporal U-Transformer (TUT) for action segmentation in videos, addressing limitations of temporal convolutions by incorporating temporal sampling and a boundary-aware loss.

Contribution

The paper proposes a novel Transformer architecture for action segmentation that eliminates temporal convolutions and introduces a boundary-aware loss to improve boundary recognition.

Findings

TUT outperforms convolution-based models on benchmark datasets.

The boundary-aware loss improves boundary detection accuracy.

Transformer-based approach reduces model complexity while maintaining performance.

Abstract

Action classification has made great progress, but segmenting and recognizing actions from long untrimmed videos remains a challenging problem. Most state-of-the-art methods focus on designing temporal convolution-based models, but the inflexibility of temporal convolutions and the difficulties in modeling long-term temporal dependencies restrict the potential of these models. Transformer-based models with adaptable and sequence modeling capabilities have recently been used in various tasks. However, the lack of inductive bias and the inefficiency of handling long video sequences limit the application of Transformer in action segmentation. In this paper, we design a pure Transformer-based model without temporal convolutions by incorporating temporal sampling, called Temporal U-Transformer (TUT). The U-Transformer architecture reduces complexity while introducing an inductive bias that adjacent frames are more likely to belong to the same class, but the introduction of coarse resolutions results in the misclassification of boundaries. We observe that the similarity distribution between a boundary frame and its neighboring frames depends on whether the boundary frame is the start or end of an action segment. Therefore, we further propose a boundary-aware loss based on the distribution of similarity scores between frames from attention modules to enhance the ability to recognize boundaries. Extensive experiments show the effectiveness of our model.