Research map / 3D vision / Face and shape
3D Shape Variational Autoencoder Latent Disentanglement via Mini-Batch Feature Swapping for Bodies and Faces
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 mini-batch feature swapping with latent consistency loss instead of an unregularized VAE | 3D mesh VAE · Face test set | Mean Rec. (mm): 0.61 -> 0.73 | Mixed |
| Use mini-batch feature swapping with latent consistency loss instead of an unregularized VAE | 3D mesh VAE · Face test set | 1-NNA CD (Δ%): 1.17 -> 0.50 | Helped |
| Use mini-batch feature swapping with latent consistency loss instead of Factor VAE | 3D mesh VAE · Face test set | Diversity: 10.51 -> 4.23 | Hurt |
| Use mini-batch feature swapping with latent consistency loss instead of DIP-VAE-I | 3D mesh VAE · Face test set | Mean Rec. (mm): 4.65 -> 0.73 | Helped |
| Train with controlled feature swaps and consistency constraints across matching and non-matching features | 3D mesh VAE · Face and body meshes | latent disentanglement | Helped |
| Use β=1e^{-2} rather than β=1e^{-4} | VAE · Face test set | latent disentanglement | Hurt |
Closest papers by meaning
- Variational Autoencoders for Deforming 3D Mesh Models (CVPR 2018 · similarity 0.89)
- Disentangling Disentanglement in Variational Autoencoders (ICML 2019 · similarity 0.89)
- Guided Variational Autoencoder for Disentanglement Learning (CVPR 2020 · similarity 0.88)
- Local Disentanglement in Variational Auto-Encoders Using Jacobian $L_1$ Regularization (NeurIPS 2021 · similarity 0.88)
- Explicit Disentanglement of Appearance and Perspective in Generative Models (NeurIPS 2019 · similarity 0.88)
- Dora: Sampling and Benchmarking for 3D Shape Variational Auto-Encoders (CVPR 2025 · similarity 0.85)
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.