Sony unveils Walrus, its most capable SpeechLLM with SOTA recognition, translation and understanding
mittu1204 · x · 2026-11-16
Sony AI has released Walrus, described as the company's most capable SpeechLLM to date. The model reportedly achieves state-of-the-art results across speech recognition, translation, and understanding tasks, with the paper/project link shared by team member AlkisKoudounas.
More from Models
- Fable 5.2 Invents 'Hard Way Day' in One-Shot Utopian AI Story — realSharonZhou · 2026-11-16
- LLMs Keep Depicting Themselves as a Floating Globe Following You Around — realSharonZhou · 2026-11-16
- Meta FAIR & NYU train 200+ MLLMs to show encoder pretraining is optional, vision-only params aren't — trungh98 · 2026-11-16
- Model Trees Launches Revamped Open-Weight Model Discovery Page — burny_tech · 2026-11-16
- Distilled image models good enough for hobbyists: 6x faster, only hands and fine text lag — Aaron_rowe81 · 2026-11-16
- Models keep getting smarter while users get babysat: a complaint about overprotective guardrails — VraserX · 2026-11-16