Research map / Label-efficient and robust learning / Recognition and long-tail
Evaluating the Disentanglement of Deep Generative Models through Manifold Topology
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 both Wasserstein RLTs and Wasserstein distance instead of the Geometry Score setup (ablation) | Topological similarity metric · dSprites real images | difference ratio: 1.60x -> 2.93x | Helped |
| Use Wasserstein RLTs instead of Euclidean-mean RLTs (ablation) | Topological similarity metric · dSprites real images | difference ratio: 1.60x -> 1.75x | Helped |
| Use Wasserstein distance instead of Euclidean distance (ablation) | Topological similarity metric · dSprites real images | difference ratio: 1.60x -> 2.14x | Helped |
| Use β-TCVAE rather than FactorVAE | Disentanglement model · dSprites | μ: 8.66 ± 1.83 -> 14.92 ± 3.46 | Helped |
| Use StyleGAN rather than ProGAN | Generative model · CelebA-HQ | μ: 0.68 ± 0.08 -> 1.03 ± 0.24 | Helped |
| Use BEGAN rather than β-VAE_B | Generative model · CelebA | μ_sup: 0.29 ± 0.25 vs 0.22 ± 0.10 | No clear effect |
Closest papers by meaning
- Learning disentangled representations via product manifold projection (ICML 2021 · similarity 0.91)
- Benchmarks, Algorithms, and Metrics for Hierarchical Disentanglement (ICML 2021 · similarity 0.90)
- Disentanglement Learning via Topology (ICML 2024 · similarity 0.90)
- What Secrets Do Your Manifolds Hold? Understanding the Local Geometry of Generative Models (ICLR 2025 · similarity 0.87)
- Transferring disentangled representations: bridging the gap between synthetic and real images (NeurIPS 2024 · similarity 0.87)
- Disentangled Representation Learning with the Gromov-Monge Gap (ICLR 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.