1 paper · 1 filter
Peixin Chen, Ying Fang, Zhifu Gao +19
Large audio language models (LALMs) have made substantial progress in audio understanding. However, autoregressive text generation introduces computational overhead and latency in…