Self-Supervised Learning for Videos: A Survey

TL;DR

This survey reviews self-supervised learning methods for videos, highlighting their categories, datasets, evaluation tasks, limitations, and future directions, emphasizing the challenges and opportunities unique to the video domain.

Contribution

It categorizes existing self-supervised video learning approaches, providing a comprehensive overview and insights into current limitations and future research directions.

Findings

Self-supervised learning reduces dependence on annotated data for videos.

Four main categories of methods: pretext tasks, generative, contrastive, cross-modal.

Identifies key datasets and evaluation benchmarks for video self-supervised learning.

Abstract

The remarkable success of deep learning in various domains relies on the availability of large-scale annotated datasets. However, obtaining annotations is expensive and requires great effort, which is especially challenging for videos. Moreover, the use of human-generated annotations leads to models with biased learning and poor domain generalization and robustness. As an alternative, self-supervised learning provides a way for representation learning which does not require annotations and has shown promise in both image and video domains. Different from the image domain, learning video representations are more challenging due to the temporal dimension, bringing in motion and other environmental dynamics. This also provides opportunities for video-exclusive ideas that advance self-supervised learning in the video and multimodal domain. In this survey, we provide a review of existing approaches on self-supervised learning focusing on the video domain. We summarize these methods into four different categories based on their learning objectives: 1) pretext tasks, 2) generative learning, 3) contrastive learning, and 4) cross-modal agreement. We further introduce the commonly used datasets, downstream evaluation tasks, insights into the limitations of existing works, and the potential future directions in this area.