AI RESEARCH
Beyond Transcription: Unified Audio Schema for Perception-Aware AudioLLMs
arXiv CS.CL
•
ArXi:2604.12506v1 Announce Type: new Recent Audio Large Language Models (AudioLLMs) exhibit a striking performance inversion: while excelling at complex reasoning tasks, they consistently underperform on fine-grained acoustic perception. We attribute this gap to a fundamental limitation of ASR-centric