Watch, read and lookup: learning to spot signs from multiple supervisors

TL;DR

This paper presents a unified learning framework for sign spotting in videos, leveraging multiple supervision sources including subtitles and dictionaries, validated on low-shot benchmarks and supported by a new BSL dictionary dataset.

Contribution

It introduces a novel integrated approach combining various supervision types for sign spotting and provides a new BSL dictionary dataset for research.

Findings

Effective sign spotting achieved on low-shot benchmarks.

Unified framework improves sign detection accuracy.

BSLDict dataset facilitates future research in sign language recognition.

Abstract

The focus of this work is sign spotting - given a video of an isolated sign, our task is to identify whether and where it has been signed in a continuous, co-articulated sign language video. To achieve this sign spotting task, we train a model using multiple types of available supervision by: (1) watching existing sparsely labelled footage; (2) reading associated subtitles (readily available translations of the signed content) which provide additional weak-supervision; (3) looking up words (for which no co-articulated labelled examples are available) in visual sign language dictionaries to enable novel sign spotting. These three tasks are integrated into a unified learning framework using the principles of Noise Contrastive Estimation and Multiple Instance Learning. We validate the effectiveness of our approach on low-shot sign spotting benchmarks. In addition, we contribute a machine-readable British Sign Language (BSL) dictionary dataset of isolated signs, BSLDict, to facilitate study of this task. The dataset, models and code are available at our project page.