FaceRefine Plus v0.2.0 speeds up MiniMax H3 video face refining by 2.7x
AnniCodex · reddit · 2026-11-23
ComfyUI-H3-FaceRefine-Plus v0.2.0 is out, an add-on for ComfyUI-H3-FaceRefine (MiniMax H3 video face refiner) fixing face wobble, pasted-on look and color drift. Heavy steps now run on GPU for a 2.7x speedup (107-frame example on RTX 5090: 244s → 90s, output within 1/255). A single Mode dropdown picks recipes: V1 (default) redraws face crops with prompt + 3 reference images for best identity; V2 uses SAM3 masks and is 1.4x faster with more wobble, plus optional VOSR2 sharpening pass. Default strength raised from 0.15 to 0.30. Available via Comfy Registry/Manager or git; example workflow included.
More from Multimodal
- Dev builds interactive 3D human eye with Gemini Neon to teach how vision works — Scobleizer · 2026-11-23
- Princeton study: diffusion models barely extrapolate basic visual skills, ~50% on counts — NielsRogge · 2026-11-23
- AI drops Goku into real life with convincing physics and lighting, blurring what's real — heypearlai · 2026-11-23
- Anima Artifacts demos a 'living magazine' you converse with instead of read — alexcovo_eth · 2026-11-23
- Turning real F1 telemetry into an AI video: 5 nights of failed experiments — BittiAI · 2026-11-23
- Full open-source pipeline for realistic 4K video: 3.7MP base, tiled latent pass, DLSS 5 — Grinderius · 2026-11-23