Research map / Image generation / Editing and style transfer
SketchDeco: Training-Free Latent Composition for Precise Sketch Colourisation
Chaitat Utintu, Yi-Zhe Song
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 |
|---|---|---|---|
| increase palette size from 1 to 5 colours (ablation) | SketchDeco global · AFHQ-cat | DCCW: 14.23 -> 9.21 | Improved |
| increase palette size from 4 to 5 colours (ablation) | SketchDeco global · AFHQ-cat | LPIPS: 0.671 -> 0.682 | Hurt |
| increase palette size from 1 to 4 colours (ablation) | SketchDeco global · PascalVOC2012 | DCCW: 15.06 -> 11.12 | Improved |
| use exceptional prompt instead of null prompt for inversion (ablation) | SketchDeco local · qualitative | colour richness / content fidelity: qualitative (Figure 10) | Improved |
| inject more attention maps (A-E combinations) (ablation) | SketchDeco local · qualitative | quality: qualitative (Figure 10) | No clear effect |
| increase scaling factor tau from 0 to 1 (ablation) | SketchDeco local · qualitative | harmonisation vs faithfulness: optimal tau in [0.3, 0.6]; tau = 0.4 used | Mixed |
| use SketchDeco instead of IDeepColor for colour palette match | T2I-Adapter global · AFHQ-cat | DCCW: 6.93 -> 10.30 | Hurt |
| use SketchDeco instead of T2I-Adapter | T2I-Adapter · AFHQ-cat | FID: 68.95 -> 50.31 | Improved |
| use SketchDeco instead of best local baseline | MangaNinja · Place365 Indoor | FID: 134.57 -> 123.87 | Improved |
| use SketchDeco instead of ColorFlow | ColorFlow · Danbooru2023 | FID: 94.87 -> 134.54 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MangaNinja — MangaNinja: Line Art Colorization with Precise Reference Following (CVPR 2025) · better on 4 of 4 shared results
- T2I-Adapter — T2I-Adapter: Learning Adapters to Dig Out More Controllable Ability for Text-to-Image Diffusion Models (AAAI 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Image Referenced Sketch Colorization Based on Animation Creation Workflow (CVPR 2025 · similarity 0.89)
- MangaNinja: Line Art Colorization with Precise Reference Following (CVPR 2025 · similarity 0.87)
- Sketch Down the FLOPs: Towards Efficient Networks for Human Sketch (CVPR 2025 · similarity 0.86)
- It's All About Your Sketch: Democratising Sketch Control in Diffusion Models (CVPR 2024 · similarity 0.86)
- One-Shot Reference-based Structure-Aware Image to Sketch Synthesis (AAAI 2025 · similarity 0.86)
- Automatic Controllable Colorization via Imagination (CVPR 2024 · similarity 0.86)
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.
