1 paper · 1 filter
Nishit Anand, Jiaqi Su, Ke Chen +5
Recent advances in Audio LLMs have achieved human-level speech recognition, yet existing systems struggle to capture paralinguistic aspects such as speaker traits, expressive varia…