Research map / Vision-language and multimodal / Human and embodied modeling
L-CoDer: Language-Based Colorization with Color-Object Decoupling Transformer
Zheng Chang, Shuchen Weng, Yu Li, Si Li, Boxin Shi
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 color-object decoupling in the tokens encoder (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | PSNR: 25.014 -> 25.504 | Improved |
| add color-object decoupling in the tokens encoder (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | LPIPS: 0.173 -> 0.159 | Improved |
| add color condition evolution with image features (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | PSNR: 25.141 -> 25.504 | Improved |
| add color condition evolution with image features (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | SSIM: 91.155% -> 91.963% | Improved |
| add bidirectional interaction between tokens (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | PSNR: 25.135 -> 25.504 | Improved |
| add bidirectional interaction between tokens (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | LPIPS: 0.168 -> 0.159 | Improved |
| add upsampling layers for token-to-pixel (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | PSNR: 25.305 -> 25.504 | Improved |
| add upsampling layers for token-to-pixel (ablation) | L-CoDer · Extended COCO-Stuff (224x224) language-based colorization | SSIM: 91.464% -> 91.963% | Improved |
| replace L-CoDe with L-CoDer transformer design | L-CoDe · Extended COCO-Stuff (224x224) language-based colorization | PSNR: 24.965 -> 25.504 | Improved |
| replace L-CoDe with L-CoDer | L-CoDe · User study reality | user preference (%): 25.76% -> 37.28% | Improved |
Closest papers by meaning
- L-CoDe:Language-Based Colorization Using Color-Object Decoupled Conditions (AAAI 2022 · similarity 0.93)
- L-CAD: Language-based Colorization with Any-level Descriptions using Diffusion Priors (NeurIPS 2023 · similarity 0.91)
- L-CoIns: Language-Based Colorization With Instance Awareness (CVPR 2023 · similarity 0.91)
- CT2: Colorization Transformer via Color Tokens (ECCV 2022 · similarity 0.88)
- Recolour What Matters: Region-Aware Colour Editing via Token-Level Diffusion (ECCV 2026 · similarity 0.87)
- ColorFormer: Image Colorization via Color Memory Assisted Hybrid-Attention Transformer (ECCV 2022 · similarity 0.87)
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.
