Online pre-training with long-form videos

TL;DR

This paper explores online pre-training methods using long-form videos, comparing masked image modeling, contrastive learning, and knowledge distillation, and finds contrastive learning most effective for action recognition tasks.

Contribution

It introduces an evaluation of multiple online pre-training techniques on long videos and demonstrates the effectiveness of contrastive learning for downstream action recognition.

Findings

Contrastive learning outperforms other methods in downstream tasks.

Learning from long-form videos benefits short video action recognition.

Online pre-training improves model performance on action recognition.

Abstract

In this study, we investigate the impact of online pre-training with continuous video clips. We will examine three methods for pre-training (masked image modeling, contrastive learning, and knowledge distillation), and assess the performance on downstream action recognition tasks. As a result, online pre-training with contrast learning showed the highest performance in downstream tasks. Our findings suggest that learning from long-form videos can be helpful for action recognition with short videos.