Gnani AI Launches Prisma v3.0 ASR for Code-Switched Indian Calls
CurieuxExplorer · x · 2026-10-28
Gnani AI introduced Gnani Prisma v3.0, a speech recognition model built for real-world Indian calls: it handles sentences that mix Hindi, English and Kannada, performs over busy streets and weak connections, and recognizes brand names the model was never trained on.
More from Multimodal
- Hand-drawn fantasy map enhanced with AI textures: original work or genAI? — SamuelBallhead · 2026-10-28
- Hi3D v1.0 image-to-3D model trends on Hugging Face — hi3d-official · 2026-10-28
- Live AI drummer JULM jams with guitarist at Barcelona AI Music Meetup Nov 27 — jordiponsdotme · 2026-10-28
- Two free deep-dives: multimodal RAG bottlenecks in order and shipping agents in a real product — ubiquitous_tech · 2026-10-28
- Image or 3DGS? Creator's LichtFeld Gaussian splatting clip is too realistic to tell — janusch_patas · 2026-10-28
- splat-repair LoRA updated: fixing Gaussian splat artifacts with Qwen Image Edit — Spy4killer · 2026-10-28