Research map / Image generation / Image restoration
Makeup-Go: Blind Reversion of Portrait Edit
Ying-Cong Chen, Xiaoyong Shen, Jiaya Jia
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 component branches from 1 to 7 (ablation) | CRN · Portrait reversion | PSNR: 36.7 -> 41.9 | Improved |
| increase component branches from 1 to 7 (ablation) | CRN · Portrait reversion | ARI: 12.5 -> 60.5 | Improved |
| use 2 branches instead of 1 (ablation) | CRN · Portrait reversion | ARI: 76.8% | Improved |
| use 7 branches instead of 6 (ablation) | CRN · Portrait reversion | ARI: 4.32% | Improved |
| add discrepancy map learning (MT) (ablation) | CRN · Meitu (MT) | PSNR: 39.8 -> 40.9 | Improved |
| add discrepancy map learning (PS) (ablation) | CRN · Photoshop Express (PS) | PSNR: 32.9 -> 33.3 | Improved |
| add discrepancy map learning (INS) (ablation) | CRN · Instagram (INS) | ARI: 3.5 -> 12.9 | Improved |
| add scale normalization (MT) (ablation) | CRN · Meitu (MT) | PSNR: 37.4 -> 40.9 | Improved |
| add scale normalization (INS) (ablation) | CRN · Instagram (INS) | ARI: -25.7 -> 12.9 | Improved |
| replace VDSR Euclidean regression with component regression (MT) | CRN vs VDSR · Meitu (MT) | ARI: 12.8 -> 60.5 | Improved |
| replace VDSR with component regression (INS) | CRN vs VDSR · Instagram (INS) | PSNR: 33.4 -> 35.9 | Improved |
| replace perception loss with component regression (PS) | CRN vs perception loss · Photoshop Express (PS) | PSNR: 22.4 -> 33.3 | Improved |
Closest papers by meaning
- BeautyGlow: On-Demand Makeup Transfer Framework With Reversible Generative Network (CVPR 2019 · similarity 0.87)
- RetouchFormer: Semi-supervised High-Quality Face Retouching Transformer with Prior-Based Selective Self-Attention (AAAI 2024 · similarity 0.86)
- Blemish-Aware and Progressive Face Retouching With Limited Paired Data (CVPR 2023 · similarity 0.85)
- PHR-DIFF: Portrait Highlights Removal via Patch-aware Diffusion Model (AAAI 2025 · similarity 0.85)
- RetouchGPT: LLM-based Interactive High-Fidelity Face Retouching via Imperfection Prompting (AAAI 2025 · similarity 0.85)
- FreeUV: Ground-Truth-Free Realistic Facial UV Texture Recovery via Cross-Assembly Inference Strategy (CVPR 2025 · similarity 0.84)
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.
