Teaching Audio-Aware Large Language Models What Does Not Hear: Mitigating Hallucinations through Synthesized Negative Samples.

Chun-Yi Kuan, Hung-yi Lee

Browse the full Interspeech paper archive.