FixAnything: 20 paired videos of LoRA let a video diffusion model repair any 3D rendering
kwangmoo_yi · x · 2026-08-26
CMU's FixAnything (ECCV 2026) uses one generalist video model to repair rendering artifacts from any 3D representation:
- Approach: framed as video-to-video translation. 3DGS/NeRF/mesh/point-cloud renders each have artifacts, but a render still preserves camera trajectory and coarse layout — an effective control signal for a pretrained video diffusion model (Wan2.1-I2V-14B)
- Data efficiency: even a very sparse point cloud (e.g. from COLMAP) gives effective camera control, with minimal LoRA finetuning on 20 paired videos
- Alignment: camera poses recovered by COLMAP from outputs serve as the reward; Flow-DPO steers toward geometrically consistent results
- Anchoring: clean training views along the trajectory act as anchors, propagating appearance, lighting and structure into degraded in-between frames
More from Multimodal
- Kling 4.0 hands-on: native 30-second 1080p generation and Omni Reference tested — HeyZoyaKhan · 2026-10-15
- Kling 4.0 generates a cinematic Tarzan riding a giant wolf, hard to tell it's AI — HeyZoyaKhan · 2026-10-15
- Kling 4.0 pitches director-style AI video: choreographed action and world-building, not random clips — alex_verem · 2026-10-15
- Adding motion graphics with just a prompt: the Claude Motion x ElevenLabs MCP workflow — CurieuxExplorer · 2026-10-15
- Google turns dense academic PDFs into animated storybooks with Gemini, Nano Banana and Veo — CurieuxExplorer · 2026-10-15
- One product photo plus one Seedance 2.5 prompt yields a full 15-second commercial — CurieuxExplorer · 2026-10-15