Spatiotemporal Contrastive Video Representation Learning

TL;DR

This paper introduces a self-supervised contrastive learning method for video representations, emphasizing the importance of spatial and temporal augmentations, achieving state-of-the-art results on Kinetics-600.

Contribution

It proposes novel spatial and temporal augmentation techniques for video contrastive learning and demonstrates significant performance improvements over prior methods.

Findings

Achieves 70.4% top-1 accuracy on Kinetics-600 with R3D-50 backbone.

Outperforms ImageNet supervised pre-training by 15.7%.

Further improves to 72.9% accuracy with R3D-152 backbone.

Abstract

We present a self-supervised Contrastive Video Representation Learning (CVRL) method to learn spatiotemporal visual representations from unlabeled videos. Our representations are learned using a contrastive loss, where two augmented clips from the same short video are pulled together in the embedding space, while clips from different videos are pushed away. We study what makes for good data augmentations for video self-supervised learning and find that both spatial and temporal information are crucial. We carefully design data augmentations involving spatial and temporal cues. Concretely, we propose a temporally consistent spatial augmentation method to impose strong spatial augmentations on each frame of the video while maintaining the temporal consistency across frames. We also propose a sampling-based temporal augmentation method to avoid overly enforcing invariance on clips that are distant in time. On Kinetics-600, a linear classifier trained on the representations learned by CVRL achieves 70.4% top-1 accuracy with a 3D-ResNet-50 (R3D-50) backbone, outperforming ImageNet supervised pre-training by 15.7% and SimCLR unsupervised pre-training by 18.8% using the same inflated R3D-50. The performance of CVRL can be further improved to 72.9% with a larger R3D-152 (2x filters) backbone, significantly closing the gap between unsupervised and supervised video representation learning. Our code and models will be available at https://github.com/tensorflow/models/tree/master/official/.