1 paper
Zhifei Xie, Zihang Liu, Ze An +8
Audio is continuous and interactive, yet most Large Audio Language Models (LALMs) remain offline and streaming systems usually specialize in ASR or spoken dialogue. We formalize th…