Research map / Image generation / GANs
Lifespan Age Transformation Synthesis
Roy Or-El, Soumyadip Sengupta, Ohad Fried, Eli Shechtman, Ira Kemelmacher-Shlizerman
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 |
|---|---|---|---|
| use the proposed model instead of IPCGAN (CACD) | Lifespan age transformation · User study vs IPCGAN CACD | Overall better (count): 27 -> 120 | Improved |
| use the proposed model instead of IPCGAN (CACD) for age accuracy | Lifespan age transformation · User study vs IPCGAN CACD | Age difference (years): 22.6 -> 11.3 | Improved |
| use the proposed model instead of IPCGAN (CACD) for identity | Lifespan age transformation · User study vs IPCGAN CACD | Same identity (count): 150 -> 136 | Hurt |
| use the proposed model instead of PyGAN | Lifespan age transformation · User study vs PyGAN (50-69) | Overall better (count): 4 -> 16 | Improved |
| use the proposed model instead of PyGAN for age accuracy | Lifespan age transformation · User study vs PyGAN (50-69) | Age difference (years): 23.1 -> 6.9 | Improved |
| use the proposed model instead of IPCGAN retrained on FFHQ-Aging | Lifespan age transformation · User study vs retrained IPCGAN | Overall better (count): 7 -> 142 | Improved |
| use the proposed model instead of IPCGAN retrained on FFHQ-Aging for identity | Lifespan age transformation · User study vs retrained IPCGAN | Same identity (count): 120 -> 137 | Improved |
| use the proposed model instead of IPCGAN retrained on FFHQ-Aging for age accuracy | Lifespan age transformation · User study vs retrained IPCGAN | Age difference (years): 5.9 -> 6.0 | No clear effect |
| use the proposed model instead of STGAN retrained on FFHQ-Aging | Lifespan age transformation · User study vs retrained STGAN | Overall better (count): 21 -> 123 | Improved |
| use the proposed model instead of STGAN retrained on FFHQ-Aging for age accuracy | Lifespan age transformation · User study vs retrained STGAN | Age difference (years): 15.6 -> 7.8 | Improved |
| add intermediate anchor age classes (2 -> 3 -> 6 anchors) (ablation) | Lifespan age transformation · Latent interpolation (qualitative) | interpolation quality (visual) | Improved |
| represent each age class with 50 vector elements instead of 1 (ablation) | Lifespan age transformation · Latent interpolation (qualitative) | interpolation quality (visual) | Improved |
Closest papers by meaning
- Age Progression and Regression with Spatial Attention Modules (AAAI 2020 · similarity 0.93)
- Continuous Face Aging via Self-Estimated Residual Age Embedding (CVPR 2021 · similarity 0.91)
- Disentangled Lifespan Face Synthesis (ICCV 2021 · similarity 0.91)
- Learning Continuous Face Age Progression: A Pyramid of GANs (TPAMI 2019 · similarity 0.91)
- Learning Face Age Progression: A Pyramid Architecture of GANs (CVPR 2018 · similarity 0.91)
- From Cradle to Cane: A Two-Pass Framework for High-Fidelity Lifespan Face Aging (NeurIPS 2025 · 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.
