Read and Attend: Temporal Localisation in Sign Language Videos

TL;DR

This paper presents a Transformer-based approach to localize and annotate signs in continuous sign language videos using weakly-aligned subtitles, significantly advancing large-scale sign language recognition.

Contribution

It introduces a method to leverage weakly-aligned subtitles for sign localization, automatically generate annotations, and improve recognition performance on a large benchmark.

Findings

Successful sign localization in continuous videos

Automatic annotation of large sign vocabulary

Outperforms previous state-of-the-art on BSL-1K

Abstract

The objective of this work is to annotate sign instances across a broad vocabulary in continuous sign language. We train a Transformer model to ingest a continuous signing stream and output a sequence of written tokens on a large-scale collection of signing footage with weakly-aligned subtitles. We show that through this training it acquires the ability to attend to a large vocabulary of sign instances in the input sequence, enabling their localisation. Our contributions are as follows: (1) we demonstrate the ability to leverage large quantities of continuous signing videos with weakly-aligned subtitles to localise signs in continuous sign language; (2) we employ the learned attention to automatically generate hundreds of thousands of annotations for a large sign vocabulary; (3) we collect a set of 37K manually verified sign instances across a vocabulary of 950 sign classes to support our study of sign language recognition; (4) by training on the newly annotated data from our method, we outperform the prior state of the art on the BSL-1K sign language recognition benchmark.