Masking Modalities for Cross-modal Video Retrieval

TL;DR

This paper introduces a novel modality masking pre-training method for video encoders that leverages appearance, sound, and speech modalities simultaneously, improving cross-modal video retrieval performance.

Contribution

It proposes a modality masking pre-training approach that enables video encoders to learn from all available modalities, including speech, which was previously underutilized.

Findings

Outperforms existing methods on How2R, YouCook2, and Condensed Movies datasets.

Enables the encoder to process appearance, sound, and speech modalities effectively.

Demonstrates the importance of multi-modal collaboration in video understanding.

Abstract

Pre-training on large scale unlabelled datasets has shown impressive performance improvements in the fields of computer vision and natural language processing. Given the advent of large-scale instructional video datasets, a common strategy for pre-training video encoders is to use the accompanying speech as weak supervision. However, as speech is used to supervise the pre-training, it is never seen by the video encoder, which does not learn to process that modality. We address this drawback of current pre-training methods, which fail to exploit the rich cues in spoken language. Our proposal is to pre-train a video encoder using all the available video modalities as supervision, namely, appearance, sound, and transcribed speech. We mask an entire modality in the input and predict it using the other two modalities. This encourages each modality to collaborate with the others, and our video encoder learns to process appearance and audio as well as speech. We show the superior performance of our "modality masking" pre-training approach for video retrieval on the How2R, YouCook2 and Condensed Movies datasets.