A Critical Study of Automatic Evaluation in Sign Language Translation

TL;DR

This study critically evaluates the effectiveness of current text-based automatic metrics for sign language translation, highlighting their limitations and the potential of LLM-based evaluators for more accurate assessments.

Contribution

It provides a comprehensive analysis of six evaluation metrics, including LLM-based methods, under various conditions, revealing their strengths and biases in assessing SLT output quality.

Findings

Lexical overlap metrics are limited in capturing semantic quality.

LLM-based evaluators better detect semantic equivalence but show bias.

BLEU is overly sensitive to minor variations.

Abstract

Automatic evaluation metrics are crucial for advancing sign language translation (SLT). Current SLT evaluation metrics, such as BLEU and ROUGE, are only text-based, and it remains unclear to what extent text-based metrics can reliably capture the quality of SLT outputs. To address this gap, we investigate the limitations of text-based SLT evaluation metrics by analyzing six metrics, including BLEU, chrF, and ROUGE, as well as BLEURT on the one hand, and large language model (LLM)-based evaluators such as G-Eval and GEMBA zero-shot direct assessment on the other hand. Specifically, we assess the consistency and robustness of these metrics under three controlled conditions: paraphrasing, hallucinations in model outputs, and variations in sentence length. Our analysis highlights the limitations of lexical overlap metrics and demonstrates that while LLM-based evaluators better capture semantic equivalence often missed by conventional metrics, they can also exhibit bias toward LLM-paraphrased translations. Moreover, although all metrics are able to detect hallucinations, BLEU tends to be overly sensitive, whereas BLEURT and LLM-based evaluators are comparatively lenient toward subtle cases. This motivates the need for multimodal evaluation frameworks that extend beyond text-based metrics to enable a more holistic assessment of SLT outputs.