@misc{indiciaee7183f14f4ca, title = {Hypothesis Stitcher for End-to-End Speaker-attributed ASR on Long-form Multi-talker Recordings}, author = {Xuankai Chang and Naoyuki Kanda and Yashesh Gaur and Xiaofei Wang and Zhong Meng and Takuya Yoshioka}, year = {2021}, url = {https://arxiv.org/abs/2101.01853}, note = {Source identifier: 2101.01853} }