Sign and Search: Sign Search Functionality for Sign Language Lexica

TL;DR

This paper introduces a reverse sign language search system using pose estimation and similarity metrics, achieving up to 90% accuracy in retrieving matching signs without prior training.

Contribution

It presents a novel webcam-based sign search method leveraging pose estimation and compares multiple similarity techniques, demonstrating high accuracy without language-specific training.

Findings

UMAP and DTW outperform other metrics in sign retrieval accuracy.

Accuracy reaches 90% at top-10 retrieval with more sign instances.

Method works across different sign languages without training.

Abstract

Sign language lexica are a useful resource for researchers and people learning sign languages. Current implementations allow a user to search a sign either by its gloss or by selecting its primary features such as handshape and location. This study focuses on exploring a reverse search functionality where a user can sign a query sign in front of a webcam and retrieve a set of matching signs. By extracting different body joints combinations (upper body, dominant hand's arm and wrist) using the pose estimation framework OpenPose, we compare four techniques (PCA, UMAP, DTW and Euclidean distance) as distance metrics between 20 query signs, each performed by eight participants on a 1200 sign lexicon. The results show that UMAP and DTW can predict a matching sign with an 80\% and 71\% accuracy respectively at the top-20 retrieved signs using the movement of the dominant hand arm. Using DTW and adding more sign instances from other participants in the lexicon, the accuracy can be raised to 90\% at the top-10 ranking. Our results suggest that our methodology can be used with no training in any sign language lexicon regardless of its size.