Research map / Image generation / Autoregressive and latent generation
HypDAE: Hyperbolic Diffusion Autoencoders for Hierarchical Few-shot Image Generation
Lingxiao Li, Kaixuan Fan, Boqing Gong, Xiangyu Yue
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 |
|---|---|---|---|
| use hyperbolic instead of Euclidean latent space (ablation) | HypDAE · Animal Faces | FID: 20.72 -> 14.31 | Improved |
| use hyperbolic instead of Euclidean latent space (ablation) | HypDAE · Animal Faces | LPIPS: 0.7288 -> 0.7415 | Improved |
| use hyperbolic instead of Euclidean latent space (ablation) | HypDAE · Flowers | FID: 25.06 -> 23.96 | Improved |
| use hyperbolic instead of Euclidean latent space (ablation) | HypDAE · VGGFaces | FID: 6.52 -> 6.25 | Improved |
| use hyperbolic instead of Euclidean latent space (ablation) | HypDAE · NABirds | FID: 8.40 -> 7.64 | Improved |
| train with CLIP pseudo labels instead of real labels (ablation) | HypDAE · Animal Faces | FID: 14.31 -> 13.14 | Improved |
| train with CLIP pseudo labels instead of real labels (ablation) | HypDAE · Flowers | FID: 23.96 -> 24.43 | Mixed |
| reduce hyperbolic radius from 5.5 to 3.0 (ablation) | HypDAE · Animal Faces | FID: 14.31 -> 20.65 | Hurt |
| reduce hyperbolic radius from 5.5 to 3.0 (ablation) | HypDAE · Animal Faces | LPIPS: 0.7415 -> 0.8964 | Improved |
| increase hyperbolic radius from 5.5 to 6.2 (ablation) | HypDAE · Animal Faces | FID: 14.31 -> 15.18 | Hurt |
| raise stochastic encoding strength from 0.95 to 1.0 (ablation) | HypDAE · Flowers | FID: 23.96 -> 28.97 | Hurt |
| lower stochastic encoding strength from 0.95 to 0.8 (ablation) | HypDAE · Flowers | FID: 23.96 -> 26.48 | Hurt |
Closest papers by meaning
- Nested Diffusion Models Using Hierarchical Latent Priors (CVPR 2025 · similarity 0.87)
- Boosting Generative Image Modeling via Joint Image-Feature Synthesis (NeurIPS 2025 · similarity 0.87)
- Kaleido Diffusion: Improving Conditional Diffusion Models with Autoregressive Latent Modeling (NeurIPS 2024 · similarity 0.86)
- Fully Hyperbolic Convolutional Neural Networks for Computer Vision (ICLR 2024 · similarity 0.86)
- DiffuseHigh: Training-Free Progressive High-Resolution Image Synthesis Through Structure Guidance (AAAI 2025 · similarity 0.85)
- Masked Autoencoders Are Effective Tokenizers for Diffusion Models (ICML 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.
