Isolated Sign Language Recognition with Segmentation and Pose Estimation
This work addresses accessibility for American Sign Language users by improving isolated sign recognition, though it is incremental as it builds on existing methods.
The authors tackled isolated sign language recognition by proposing a model that integrates pose estimation, segmentation, and a ResNet-Transformer backbone to reduce computational costs and handle signer variability, achieving competitive results on benchmark datasets.
The recent surge in large language models has automated translations of spoken and written languages. However, these advances remain largely inaccessible to American Sign Language (ASL) users, whose language relies on complex visual cues. Isolated sign language recognition (ISLR) - the task of classifying videos of individual signs - can help bridge this gap but is currently limited by scarce per-sign data, high signer variability, and substantial computational costs. We propose a model for ISLR that reduces computational requirements while maintaining robustness to signer variation. Our approach integrates (i) a pose estimation pipeline to extract hand and face joint coordinates, (ii) a segmentation module that isolates relevant information, and (iii) a ResNet-Transformer backbone to jointly model spatial and temporal dependencies.