Research map / Probabilistic and scientific modeling / Variational and generative sampling
Guided Variational Autoencoder for Disentanglement Learning
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 |
|---|---|---|---|
| Add unsupervised guidance to VAE | VAE · 2D Shapes | Z-Diff: 78.2 -> 99.2 | Helped |
| Add unsupervised guidance to β-TCVAE | β-TCVAE · 2D Shapes | Factor: 0.6968 -> 0.7294 | Helped |
| Add unsupervised guidance to β-TCVAE | β-TCVAE · MNIST | classification error (d_z=32): 1.24%±0.05 -> 1.10%±0.03 | Helped |
| Add supervised excitation-inhibition guidance | ALI-VAE · CIFAR10 Truck-Automobile | FID: 78.92 -> 72.31 | Helped |
| Add supervised guidance to the few-shot model | Baseline · Omniglot 20-way 1-shot | classification accuracy: 91.4% -> 92.1% | Helped |
| Add lightweight guidance (ablation) | AE · MNIST | classification error: 1.37%±0.05 -> 1.46%±0.06; 1.06%±0.04 -> 1.00%±0.06; 1.34%±0.04 -> 1.10%±0.08 | Mixed |
| Remove geometric deformation from deformable PCA guidance (ablation) | Unsupervised Guided-VAE · MNIST latent traversal | geometry encoded in designated latent variables | Hurt |
| Remove adversarial inhibition (ablation) | Supervised Guided-VAE · CelebA latent traversal | attribute disentanglement | Hurt |
Closest papers by meaning
- An Identifiable Double VAE For Disentangled Representations (ICML 2021 · similarity 0.92)
- Disentangling Disentanglement in Variational Autoencoders (ICML 2019 · similarity 0.91)
- Variational Autoencoders Pursue PCA Directions (by Accident) (CVPR 2019 · similarity 0.91)
- Learning Disentangled Representations with Semi-Supervised Deep Generative Models (NeurIPS 2017 · similarity 0.91)
- $\alpha$TC-VAE: On the relationship between Disentanglement and Diversity (ICLR 2024 · similarity 0.89)
- Why do Variational Autoencoders Really Promote Disentanglement? (ICML 2024 · 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.