TY - RPRT TI - A Speech-LLM Framework for Enrollment-Free Wake-Word-Guided Target-Speaker Recognition in Complex Noisy Scenarios for Seamless Human-Machine Interaction AU - Yuxiang Mei AU - Dongxing Xu AU - Yanhua Long PY - 2026 UR - https://searcharxiv.com/abs/2611.19844 ID - 2611.19844 N1 - Source identifier: 2611.19844 ER -