Research map / Image generation / Editing and style transfer
Custom Structure Preservation in Face Aging
Guillermo Gomez-Trenado, Stéphane Lathuilière, Pablo Mesejo, Óscar Cordón
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 a separate style encoder (vs average pooling of content encoder output) (ablation) | CUSP · FFHQ-RR | LPIPS: 0.84 -> 0.78 | Improved |
| add a separate style encoder (ablation) | CUSP · FFHQ-RR | Mean KID: 0.0163 -> 0.0089 | Improved |
| add skip connections in the second-to-last layer (ablation) | CUSP · FFHQ-RR | LPIPS: 1.70 -> 0.78 | Improved |
| add skip connections in the second-to-last layer (ablation) | CUSP · FFHQ-RR | Age MAE: 6.17 -> 6.29 | No clear effect |
| use skip connections only in second-to-last layer instead of every layer (ablation) | CUSP · FFHQ-RR | LPIPS: 1.85 -> 0.78 | Improved |
| class-independent GB mask instead of top-1 class GB (ablation) | CUSP · FFHQ-RR | LPIPS: 1.25 -> 0.78 | Improved |
| class-independent GB mask instead of top-1 class GB (ablation) | CUSP · FFHQ-RR | Mean KID: 0.0145 -> 0.0089 | Improved |
| class-independent GB mask instead of top-1 class GB (ablation) | CUSP · FFHQ-RR | Age MAE: 6.19 -> 6.29 | Hurt |
| use local mask-based blur instead of global blur (ablation) | CUSP CP · FFHQ-RR test | LPIPS (x10): 1.56 -> 0.78 | Improved |
| use local mask-based blur instead of global blur (ablation) | CUSP CP · FFHQ-RR test | Age MAE: 5.84 -> 6.29 | Hurt |
| increase preservation from low to high (CUSP LP to HP) (ablation) | CUSP · FFHQ-RR test | Age MAE: 6.07 -> 9.05 | Hurt |
| CUSP HP vs best baseline LATS on target age accuracy | CUSP HP · CelebA-HQ young-to-old | predicted age: 55.33 -> 67.76 | Improved |
Closest papers by meaning
- From Cradle to Cane: A Two-Pass Framework for High-Fidelity Lifespan Face Aging (NeurIPS 2025 · similarity 0.87)
- Continuous Face Aging via Self-Estimated Residual Age Embedding (CVPR 2021 · similarity 0.86)
- Self-Supervised Selective-Guided Diffusion Model for Old-Photo Face Restoration (NeurIPS 2025 · similarity 0.85)
- CA-Edit: Causality-Aware Condition Adapter for High-Fidelity Local Facial Attribute Editing (AAAI 2025 · similarity 0.84)
- Text-Guided 3D Face Synthesis - From Generation to Editing (CVPR 2024 · similarity 0.84)
- DreamSalon: A Staged Diffusion Framework for Preserving Identity-Context in Editable Face Generation (CVPR 2024 · similarity 0.83)
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.
