Berkeley's StyleStream 2.0 cuts streaming voice conversion latency to 520ms, 2x faster
Berkeley · hf · 2026-10-23
UC Berkeley researchers released StyleStream 2.0, addressing two limitations of v1: 1s end-to-end latency on consumer hardware and the lack of a text interface for voice design. Key points:
- A few-step pixel MeanFlow model, fine-tuned with Self Forcing-inspired autoregressive feedback, cuts latency to 520ms on a consumer GPU — a 2x speedup while maintaining quality
- A unified flow matching model enables text-based voice design and editing
Paper and demo are available on Hugging Face.
More from Multimodal
- Ant's Ming-3.0-tts hits top-tier TTS quality with just 0.5B params, MIT licensed — ArtificialAnlys · 2026-10-23
- Suno Highlights Five Niche Features Most Users Don't Know About — suno · 2026-10-23
- Meshy offers Hermes users a month of Pro with API access for $1 — Teknium · 2026-10-23
- Customuse launches Smart Mesh AI with geometry-aware adaptive character clothing tool — SimplyAnnisa · 2026-10-23
- AI-generated movie poster 'The Chase' crams a wolf, a plane and muscle cars into one chase — Aggravating-Sir-115 · 2026-10-23
- SteerTTS-edge: an 11.84MB steerable TTS with unchanged WER running 19x realtime on CPU — Working_Hat5120 · 2026-10-23