Research map / Probabilistic and scientific modeling / Variational and generative sampling
Explicitly disentangling image content from translation and rotation with spatial-VAE
Tristan Bepler, Ellen Zhong, Kotaro Kelley, Edward Brignole, Bonnie Berger
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 explicitly inferred rotation and translation variables, structurally separated from content latents | VAE · rotated MNIST and rotated and translated MNIST | test ELBO | Improved |
| Add explicitly inferred rotation and translation variables, structurally separated from content latents | VAE · 5HDB | ELBO | Improved |
| Add explicitly inferred rotation and translation variables, structurally separated from content latents | VAE · 5HDB | conformation correlation | Improved |
| Add explicitly inferred rotation and translation variables, structurally separated from content latents | VAE · 5HDB | rotation correlation | Improved |
| Add rotation and translation inference instead of fixing theta and translation to zero (ablation) | spatial-VAE · MNIST, rotated MNIST, and rotated and translated MNIST | test ELBO | Improved |
| Use spatial-VAE rather than vanilla VAE | image generative modeling · Galaxy Zoo | rotation- and translation-invariant content representation | Improved |
Closest papers by meaning
- Unsupervised Object Representation Learning using Translation and Rotation Group Equivariant VAE (NeurIPS 2022 · similarity 0.90)
- EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling (ICML 2025 · similarity 0.88)
- Disentangling Factors of Variation with Cycle-Consistent Variational Auto-Encoders (ECCV 2018 · similarity 0.88)
- Variational Autoencoders Pursue PCA Directions (by Accident) (CVPR 2019 · similarity 0.88)
- Local Disentanglement in Variational Auto-Encoders Using Jacobian $L_1$ Regularization (NeurIPS 2021 · similarity 0.88)
- Disentangling Disentanglement in Variational Autoencoders (ICML 2019 · 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.
