FastDialogueSidon open-sources 100x faster two-speaker dialogue separation at matching quality
HigurashiAI · x · 2026-10-27
higurashi-ai has released FastDialogueSidon, an open-source high-speed version of the full-duplex two-speaker dialogue separation and restoration model DialogueSidon.
- Takes raw 16 kHz mono mixtures (phone/conversational audio, up to 120 s per window) and outputs two separated 24 kHz speaker tracks
- Replaces DialogueSidon's 30 diffusion steps with a single DiT step and swaps its DAC decoder for a lightweight 30M-parameter SAN-WaveNeXt2 latent decoder
- Runs on fused GPU kernels (Triton and TensorRT backends), delivering 100x speedup over the official implementation on H100 at matching intelligibility
Available on GitHub under Apache-2.0 and on Hugging Face.
Related event: FastDialogueSidon open-sources 100x faster dialogue separation(2 posts)→
More from Research
- Recursive self-improvement has no single starting moment — Cosmos Institute essay dissects the intelligence explosion — sebkrier · 2026-11-14
- Four frontier models score 97.8–100 on a self-built materials-science research eval — v_aravind · 2026-11-14
- Measuring face reconstruction fidelity with AuraFace embedding cosine similarity — ostrisai · 2026-11-14
- 3DSceneBench: coding agents rebuild livable 3D rooms but top functionality score is just 42.1 — yizhu59 · 2026-11-14
- MIT team trains open 27B model with zero demos to score 18/30 on a real robot it never touched — songhan_mit · 2026-11-14
- New paper argues alignment is a system-level context property, not just model weights — rayanpal_ · 2026-11-14