analiza aiqbench
This research introduces Anuvani-Diarization, a zero-shot speaker diarization system optimized for India's 22 scheduled languages using self-supervised models and uncertainty-aware supervision. The study demonstrates significant performance gains over existing systems while highlighting critical annotation convention biases that impact error rates across diverse linguistic contexts.
Anuvani-Diarization: Zero Shot Speaker Diarization and Annotation Conventions Across 22 Languages
arXiv:2610.20327v1 Announce Type: new Abstract: India's 22 scheduled languages are spoken by over a billion people in conversations that mix languages, often within one turn, overlap freely, and take place over telephones, in meetings, and in media. Knowing who spoke when, the task of speaker diarization, is essential to speech applications for such conversation: voice agents conversing with several people at once, speaker attributed transcription of meetings and calls, contact center analytics, and media monitoring. We present a systematic study of speaker diarization across all 22 languages on IndicDiarBench and introduce Anuvani-Diarization, which adapts a self-supervised end-to-end segmentation model to Indian conversation with uncertainty aware silver supervision. Evaluated zero shot alongside nine recent diarization systems, it reaches a duration weighted diarization error rate (DER) of 16.78% on all 1,164 recordings, the lowest of all systems we ran and second only to the best published result (16.0%), and leads the three strongest rivals in 14 of 22 languages. It matches its parent on six international benchmarks. We also show that the benchmark's utterance level annotation convention is a hidden variable worth 1.9 to 4.1 DER points, depending on the system, and that a convention learned from other Indian transcriptions does not transfer.
Abstract:India's 22 scheduled languages are spoken by over a billion people in conversations that mix languages, often within one turn, overlap freely, and take place over telephones, in meetings, and in media. Knowing who spoke when, the task of speaker diarization, is essential to speech applications for such conversation: voice agents conversing with several people at once, speaker attributed transcription of meetings and calls, contact center analytics, and media monitoring. We present a systematic study of speaker diarization across all 22 languages on IndicDiarBench and introduce Anuvani-Diarization, which adapts a self-supervised end-to-end segmentation model to Indian conversation with uncertainty aware silver supervision. Evaluated zero shot alongside nine recent diarization systems, it reaches a duration weighted diarization error rate (DER) of 16.78% on all 1,164 recordings, the lowest of all systems we ran and second only to the best published result (16.0%), and leads the three strongest rivals in 14 of 22 languages. It matches its parent on six international benchmarks. We also show that the benchmark's utterance level annotation convention is a hidden variable worth 1.9 to 4.1 DER points, depending on the system, and that a convention learned from other Indian transcriptions does not transfer.
Computation and Language (cs.CL)
arXiv:2610.20327 [cs.CL]
(or arXiv:2610.20327v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.20327 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 291 cuvinte
Subiecte conexe
CARDAMOM: A Micro-Dialectal Arabic Speech Dataset for ASR
arXiv · 30 sept. 2026
Measuring Negative Campaigning across Languages with Large Language Models: A Study of 18 Million Tweets in 19 Countries
arXiv · 28 iul. 2026
Zero-Shot Multi-Disease Labeling of Chest, Abdomen, and Pelvis CT Reports Using Open-Weight Large Language Models: The Effect of Labeling Conventions
arXiv · 7 aug. 2026
Speech Generation Speaker Poisoning: Capability Erasure in Zero-Shot Text-to-Speech
arXiv · 16 sept. 2026