LumiTokens: 3D Relighting via Token-Space Lighting Transformation
qixing_huang · x · 2026-08-31
This ECCV 2026 paper introduces LumiTokens, a framework that formulates 3D relighting as a direct transformation on latent scene tokens, bypassing the need for explicit geometry, material decomposition, or rendering equations. Key features include:
- Scene Token Editor: Processes scene and light-ray tokens via self-attention to output updated tokens decodable into multi-view-consistent images.
- Progressive Relighting: Allows incremental lighting additions (environment, point, area lights) that compose directly in token space without re-encoding.
The method achieves competitive quality with sparse views and avoids per-scene optimization.
More from Multimodal
- Pollo AI launches Recast video restyling, 60% off for launch — SimplyAnnisa · 2026-10-16
- Tripo hosts first world-building hackathon in SF with World Labs and Runway talks — mhdfaran · 2026-10-16
- Kling 4.0 boosts text-to-video with character consistency and lip sync — CurieuxExplorer · 2026-10-16
- 5.8M-param Speech Event Detector adds laughs, breaths and背景 sounds to Whisper transcripts — Ayush Kumar Bar · 2026-10-16
- Huawei's SUPRA-Tex turns one image into relightable 3D materials with WYSIWYG fidelity — Chen Yang · 2026-10-16
- Giant Michael Jackson AI fashion images show the power of extreme proportions — OVolosin82152 · 2026-10-16