Exploring Scalability of Self-Training for Open-Vocabulary Temporal Action Localization

TL;DR

This paper investigates the use of self-training with unlabeled web videos to improve open-vocabulary temporal action localization, demonstrating enhanced generalizability and proposing a new evaluation benchmark.

Contribution

It introduces a scalable self-training framework for OV-TAL using pseudo-labels from unlabeled videos and proposes a new benchmark for comprehensive evaluation.

Findings

Self-training with web videos improves OV-TAL performance.

The new benchmark reveals limitations of existing evaluation schemes.

Gemini-1.5 achieves competitive results on the new benchmark.

Abstract

The vocabulary size in temporal action localization (TAL) is limited by the scarcity of large-scale annotated datasets. To overcome this, recent works integrate vision-language models (VLMs), such as CLIP, for open-vocabulary TAL (OV-TAL). However, despite the success of VLMs trained on extensive datasets, existing OV-TAL methods still rely on human-labeled TAL datasets of limited size to train action localizers, limiting their generalizability. In this paper, we explore the scalability of self-training with unlabeled YouTube videos for OV-TAL. Our approach consists of two stages: (1) a class-agnostic action localizer is trained on a human-labeled TAL dataset to generate pseudo-labels for unlabeled videos, and (2) the large-scale pseudo-labeled dataset is then used to train the localizer. Extensive experiments demonstrate that leveraging web-scale videos in self-training significantly enhances the generalizability of an action localizer. Additionally, we identify limitations in existing OV-TAL evaluation schemes and propose a new benchmark for thorough assessment. Finally, we showcase the TAL performance of the large multimodal model Gemini-1.5 on our new benchmark. Code is released at https://github.com/HYUNJS/STOV-TAL.