This paper provides a detailed description of the Hitachi-JHU system that was submitted to the Third DIHARD Speech Diarization Challenge. The system outputs the ensemble results of the five subsystems: two x-vector-based subsystems, two end-to-end neural diarization-based subsystems, and one hybrid subsystem. We refine each system and all five subsystems become competitive and complementary. After the DOVER-Lap based system combination, it achieved diarization error rates of 11.58 % and 14.09 % in Track 1 full and core, and 16.94 % and 20.01 % in Track 2 full and core, respectively. With their results, we won second place in all the tasks of the challenge.
@article{arxiv.2102.01363,
title = {The Hitachi-JHU DIHARD III System: Competitive End-to-End Neural Diarization and X-Vector Clustering Systems Combined by DOVER-Lap},
author = {Shota Horiguchi and Nelson Yalta and Paola Garcia and Yuki Takashima and Yawen Xue and Desh Raj and Zili Huang and Yusuke Fujita and Shinji Watanabe and Sanjeev Khudanpur},
journal= {arXiv preprint arXiv:2102.01363},
year = {2021}
}