Research map / Image generation / Editing and style transfer
Hierarchical Face Aging through Disentangled Latent Characteristics
Peipei Li, Huaibo Huang, Yibo Hu, Xiang Wu, Ran He, Zhenan Sun
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 disentangled adversarial loss L_adv (ablation) | DAAE · Morph face verification (AG1/AG2/AG3) | verification rate (%): 95.74 -> 99.48 (AG1), 95.61 -> 99.36 (AG2), 95.60 -> 99.36 (AG3) | Improved |
| add age regularization L_reg^age (ablation) | DAAE · Morph face verification (AG1/AG2/AG3) | verification rate (%): 97.84 -> 99.48 (AG1), 97.89 -> 99.36 (AG2), 97.89 -> 99.36 (AG3) | Improved |
| add identity regularization L_reg^id (ablation) | DAAE · Morph face verification (AG1/AG2/AG3) | verification rate (%): 96.21 -> 99.48 (AG1), 93.01 -> 99.36 (AG2), 93.09 -> 99.36 (AG3) | Improved |
| hierarchical generator: identity into early layers and age into late layers instead of the reverse (Setting I) (ablation) | DAAE · Morph face verification (AG1/AG2/AG3) | verification rate (%): 89.04 -> 99.48 (AG1), 87.35 -> 99.36 (AG2), 87.35 -> 99.36 (AG3) | Improved |
| hierarchical generator instead of concatenating age, identity, extraneous at input layer (Setting II) (ablation) | DAAE · Morph face verification (AG1/AG2/AG3) | verification rate (%): 76.10 -> 99.48 (AG1), 72.20 -> 99.36 (AG2), 72.15 -> 99.36 (AG3) | Improved |
| use unified DAAE model instead of per-age-group models for identity preservation | DAAE · Morph face verification AG3 | verification rate (%): 91.85 -> 99.36 | Improved |
| use DAAE age posterior mean instead of prior methods for age estimation | DAAE · Morph age estimation | MAE: 2.35 -> 2.23 | Improved |
Closest papers by meaning
- TimeBooth: Disentangled Facial Invariant Representation for Diverse and Personalized Face Aging (ICCV 2025 · similarity 0.93)
- Disentangled Representation for Age-Invariant Face Recognition: A Mutual Information Minimization Perspective (ICCV 2021 · similarity 0.91)
- Age Progression/Regression by Conditional Adversarial Autoencoder (CVPR 2017 · similarity 0.90)
- Continuous Face Aging via Self-Estimated Residual Age Embedding (CVPR 2021 · similarity 0.90)
- Learning Continuous Face Age Progression: A Pyramid of GANs (TPAMI 2019 · similarity 0.90)
- From Cradle to Cane: A Two-Pass Framework for High-Fidelity Lifespan Face Aging (NeurIPS 2025 · similarity 0.88)
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.
