AutoSign: Direct Pose-to-Text Translation for Continuous Sign Language Recognition

TL;DR

AutoSign introduces a novel autoregressive transformer model that directly translates pose sequences into text for continuous sign language recognition, overcoming traditional pipeline limitations and improving accuracy.

Contribution

It proposes a direct pose-to-text translation model using a decoder-only transformer, bypassing alignment-based methods and enhancing signer-independent CSLR performance.

Findings

Achieves up to 6.1% WER improvement on Isharah-1000 dataset.

Demonstrates the effectiveness of hand and body gestures for signer-independent recognition.

Eliminates multi-stage pipelines, reducing error propagation.

Abstract

Continuously recognizing sign gestures and converting them to glosses plays a key role in bridging the gap between the hearing and hearing-impaired communities. This involves recognizing and interpreting the hands, face, and body gestures of the signer, which pose a challenge as it involves a combination of all these features. Continuous Sign Language Recognition (CSLR) methods rely on multi-stage pipelines that first extract visual features, then align variable-length sequences with target glosses using CTC or HMM-based approaches. However, these alignment-based methods suffer from error propagation across stages, overfitting, and struggle with vocabulary scalability due to the intermediate gloss representation bottleneck. To address these limitations, we propose AutoSign, an autoregressive decoder-only transformer that directly translates pose sequences to natural language text, bypassing traditional alignment mechanisms entirely. The use of this decoder-only approach allows the model to directly map between the features and the glosses without the need for CTC loss while also directly learning the textual dependencies in the glosses. Our approach incorporates a temporal compression module using 1D CNNs to efficiently process pose sequences, followed by AraGPT2, a pre-trained Arabic decoder, to generate text (glosses). Through comprehensive ablation studies, we demonstrate that hand and body gestures provide the most discriminative features for signer-independent CSLR. By eliminating the multi-stage pipeline, AutoSign achieves substantial improvements on the Isharah-1000 dataset, achieving an improvement of up to 6.1\% in WER score compared to the best existing method.