Measuring Fairness in Large Audio Language Models via Semantic-Aware Bias Estimation

Zhe Liu

Large Audio Language Models (LALMs) have seen increasing use for audio understanding tasks such as speech recognition and audio question answering, raising concerns about fairness…