Research map / 3D vision / 3D generation
Color3D: Controllable and Consistent 3D Colorization with Personalized Colorizer
Yecong Wan, Mingwen Shao, Renlong Wu, Wangmeng Zuo
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add key view selection (ablation) | Color3D baseline · Mip-NeRF 360 (language-guided) | FID: 115.66 -> 100.28 | Improved |
| add single view augmentation (ablation) | Color3D · Mip-NeRF 360 (language-guided) | FID: 100.28 -> 85.25 | Improved |
| add fine-tuned personalized colorizer (ablation) | Color3D · Mip-NeRF 360 (language-guided) | FID: 85.25 -> 75.48 | Improved |
| add Lab Gaussian representation (ablation) | Color3D · Mip-NeRF 360 (language-guided) | FID: 75.48 -> 68.23 | Improved |
| add Lab Gaussian representation (ablation) | Color3D · Mip-NeRF 360 (language-guided) | Matching Error: 0.064 -> 0.058 | Improved |
| use one key view instead of three (ablation) | Color3D · Mip-NeRF 360 (automatic) | FID: 52.38 -> 39.03 | Improved |
| use one key view instead of two (ablation) | Color3D · Mip-NeRF 360 (automatic) | Colorful score: 27.36 -> 33.36 | Improved |
| per-scene colorizer instead of generalizable colorizer (ablation) | Color3D with AnySplat · DL3DV-140 (automatic) | FID: 82.36 -> 66.78 | Improved |
| replace 3DGS+ImageColorizer with Color3D | 3DGS pipeline · Mip-NeRF 360 (automatic) | Matching Error: 0.135 -> 0.082 | Improved |
| replace 3DGS+ImageColorizer with Color3D | 3DGS pipeline · LLFF (automatic) | Colorful score: 34.07 -> 33.99 | No clear effect |
| Color3D instead of ColorNeRF | Static scene colorization · LLFF (automatic) | FID: 58.26 -> 35.10 | Improved |
Closest papers by meaning
- Control Color: Multimodal Diffusion-Based Interactive Image Colorization (IJCV 2025 · similarity 0.88)
- Learning Generative Neural Networks for 3D Colorization (AAAI 2018 · similarity 0.88)
- Structural Consistency and Controllability for Diverse Colorization (ECCV 2018 · similarity 0.87)
- BigColor: Colorization Using a Generative Color Prior for Natural Images (ECCV 2022 · similarity 0.87)
- GLUT: 3D Gaussian Lookup Table for Continuous Color Transformation (NeurIPS 2026 · similarity 0.86)
- Automatic Controllable Colorization via Imagination (CVPR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
