Research map / Image generation / Editing and style transfer
Explicit Disentanglement of Appearance and Perspective in Generative Models
Read the paperCheck 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 conditional rather than unconditional inference (ablation) | VITAE · dSprites | D_score: 0.22 -> 0.38 | Helped |
| Use conditional rather than unconditional inference (ablation) | VITAE · MNIST | D_score: 0.782 -> 0.884 | Helped |
| Use conditional rather than unconditional inference (ablation) | VITAE · SMPL | D_score: 0.673 -> 0.943 | Helped |
| Use conditional rather than unconditional inference (ablation) | VITAE · MNIST | ELBO: -142 -> -139 | Helped |
| Use conditional rather than unconditional inference (ablation) | VITAE · dSprites | ELBO: -55.25 -> -68.26 | Hurt |
| Use conditional rather than unconditional inference (ablation) | VITAE · MNIST | log likelihood: -143 -> -141 | Helped |
| Use conditional rather than unconditional inference (ablation) | VITAE · SMPL | log likelihood: -8.55\times 10^3 -> -8.52\times 10^3 | Helped |
| Parametrize affine transformations as diffeomorphisms using the matrix exponential (ablation) | C-VITAE · MNIST | ELBO | Helped |
Closest papers by meaning
- Orthogonality-Enforced Latent Space in Autoencoders: An Approach to Learning Disentangled Representations (ICML 2023 · similarity 0.91)
- Transferring disentangled representations: bridging the gap between synthetic and real images (NeurIPS 2024 · similarity 0.90)
- Overcoming the Disentanglement vs Reconstruction Trade-off via Jacobian Supervision (ICLR 2019 · similarity 0.89)
- Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning View (ICLR 2022 · similarity 0.89)
- Demystifying Inductive Biases for (Beta-)VAE Based Architectures (ICML 2021 · similarity 0.89)
- Disentangled Representation Learning via Modular Compositional Bias (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.