Benchmark RadarAI BENCHMARK PROFILE
Indic DiarBench
Multilingual joint diarization and ASR benchmark for 22 Indian languages, with ~108 hours of human-corrected multi-speaker audio from meetings, far-field, and in-the-wild sources, including code-mixing and overlaps.
- Released
- 2026-07-26
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Provides a standardized evaluation suite for speaker diarization and ASR on Indian languages, addressing a gap in multilingual speech technology assessment.
Motivation
In this work, we introduce Indic DiarBench, a speaker diarization and ASR benchmark dataset spanning all 22 scheduled languages of India.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.