ISCA Archive Interspeech 2022
ISCA Archive Interspeech 2022

Unsupervised Speaker Diarization that is Agnostic to Language, Overlap-Aware, and Tuning Free

Md Iftekhar Tanveer, Diego Casabuena, Jussi Karlgren, Rosie Jones

Podcasts are conversational in nature and speaker changes are frequent---requiring speaker diarization for content understanding. We propose an unsupervised technique for speaker diarization without relying on language-specific components. The algorithm is overlap-aware and does not require information about the number of speakers. Our approach shows 79% improvement on purity scores (34% on F-score) against the Google Cloud Platform solution on podcast data.


doi: 10.21437/Interspeech.2022-10605

Cite as: Tanveer, M.I., Casabuena, D., Karlgren, J., Jones, R. (2022) Unsupervised Speaker Diarization that is Agnostic to Language, Overlap-Aware, and Tuning Free. Proc. Interspeech 2022, 1481-1485, doi: 10.21437/Interspeech.2022-10605

@inproceedings{tanveer22_interspeech,
  author={Md Iftekhar Tanveer and Diego Casabuena and Jussi Karlgren and Rosie Jones},
  title={{Unsupervised Speaker Diarization that is Agnostic to Language, Overlap-Aware, and Tuning Free}},
  year=2022,
  booktitle={Proc. Interspeech 2022},
  pages={1481--1485},
  doi={10.21437/Interspeech.2022-10605}
}