HEAR Who Said What: Unlocking Speaker-Attributed Reasoning via Counterfactual Voice Grounding
A2R, a 30B model optimized on Counterfactual Audio with Speaker-level Hard negatives (CASH), a dataset designed to guide the model to prioritize acoustic vocal cues over linguistic signals, achieves strong performance on HEAR and exhibits zero-shot generalization to diverse multi-speaker downstream tasks, demonstrating that learned speaker attribution unlocks the model's latent capacity for speaker-aware reasoning.
Dongwook Lee, Sangkwon Park, Eunwoo Song et al.
· 0 citations