1
G-STAR: End-to-End Global Speaker-Tracking Attributed Recognition
全球说话人跟踪与识别一体化端到端新框架，攻克多人场景下的身份与轨迹联合建模难题
arXiv:2603.10468v2 Announce Type: replace-cross Abstract: We study timestamped speaker-attributed automatic speech recognition (SA-ASR) for long-form,…