882-Hour Diarized Bengali Speech Corpus Lipi-Ghor Released with Build Pipeline
teamvillagers · hf · 2026-10-29
A team from BUET CS Fest 2026's DL Sprint releases Lipi-Ghor (Lipi-Ghor-882), an 882-hour, multi-speaker, multi-domain Bengali speech corpus — reportedly the first at this scale supporting joint ASR and speaker diarization.
- Built with weak supervision, it provides long-form annotations with speaker diarization plus a gold evaluation split for benchmarking ASR models.
- The scalable corpus-building pipeline is also open-sourced, under Apache 2.0 / CC-BY-4.0 licenses.
More from Research
- Preprint: Human peer reviewers caught only 27% of major errors, all 15 AI models beat them (53-96%) — anshulkundaje · 2026-10-29
- New paper rules out Hamiltonian Phase States as a route to quantum-only Microcrypt — StefanoGogioso · 2026-10-29
- MSR AI for Science team hiring postdoc to generate molecular data for Skala — FrankNoeBerlin · 2026-10-29
- ML-enhanced resonant soft X-ray scattering enables sub-monolayer tracking of buried-interface molecular states — carbon-lab · 2026-10-29
- Any3R: one feedforward model handles 13 camera types for 3D reconstruction, beating OmniVGGT and PanoVGGT — zhenjun_zhao · 2026-10-29
- GeoGSplat: geometry priors top DTU and NRGBD with just 65K Gaussians — zhenjun_zhao · 2026-10-29