Sign Language Video Retrieval with Free-Form Textual Queries

TL;DR

This paper introduces a novel sign language retrieval task using free-form textual queries and proposes a framework called SPOT-ALIGN to improve sign video embeddings by iterative sign spotting and feature alignment, enhancing retrieval accuracy.

Contribution

The work presents the first approach to sign language video retrieval with natural language queries and introduces SPOT-ALIGN to improve embedding quality through iterative data expansion.

Findings

SPOT-ALIGN improves sign recognition accuracy.

Enhanced embeddings lead to better video retrieval performance.

The approach effectively scales training data for sign language models.

Abstract

Systems that can efficiently search collections of sign language videos have been highlighted as a useful application of sign language technology. However, the problem of searching videos beyond individual keywords has received limited attention in the literature. To address this gap, in this work we introduce the task of sign language retrieval with free-form textual queries: given a written query (e.g., a sentence) and a large collection of sign language videos, the objective is to find the signing video in the collection that best matches the written query. We propose to tackle this task by learning cross-modal embeddings on the recently introduced large-scale How2Sign dataset of American Sign Language (ASL). We identify that a key bottleneck in the performance of the system is the quality of the sign video embedding which suffers from a scarcity of labeled training data. We, therefore, propose SPOT-ALIGN, a framework for interleaving iterative rounds of sign spotting and feature alignment to expand the scope and scale of available training data. We validate the effectiveness of SPOT-ALIGN for learning a robust sign video embedding through improvements in both sign recognition and the proposed video retrieval task.