Research map / Probabilistic and scientific modeling / Variational and generative sampling
Perceptual Generative Autoencoders
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 LPGA instead of VAE | Autoencoder-based generation · MNIST | FID: 13.83±0.06 -> 10.34±0.15 | Helped |
| Use LPGA instead of VAE | Autoencoder-based generation · CIFAR-10 | FID: 115.74±0.63 -> 55.87±0.25 | Helped |
| Use LPGA instead of VAE | Autoencoder-based generation · CelebA | FID: 43.60±0.33 -> 14.53±0.52 | Helped |
| Use VPGA instead of VAE | Autoencoder-based generation · MNIST | FID: 13.83±0.06 -> 4.97±0.07 | Helped |
| Use VPGA instead of VAE | Autoencoder-based generation · CIFAR-10 | FID: 115.74±0.63 -> 51.51±1.16 | Helped |
| Use VPGA instead of VAE | Autoencoder-based generation · CelebA | FID: 43.60±0.33 -> 24.73±1.25 | Helped |
| Use LVPGA instead of VAE | Autoencoder-based generation · MNIST | FID: 13.83±0.06 -> 6.32±0.16 | Helped |
| Use LVPGA instead of VAE | Autoencoder-based generation · CIFAR-10 | FID: 115.74±0.63 -> 52.94±0.89 | Helped |
| Use LVPGA instead of VAE | Autoencoder-based generation · CelebA | FID: 43.60±0.33 -> 13.80±0.20 | Helped |
| Add the latent reconstruction loss (ablation) | LPGA · Image datasets in the ablation study | FID | Helped |
Closest papers by meaning
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.90)
- InfoVAE: Balancing Learning and Inference in Variational Autoencoders (AAAI 2019 · similarity 0.89)
- EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling (ICML 2025 · similarity 0.89)
- Complexity Matters: Rethinking the Latent Space for Generative Modeling (NeurIPS 2023 · similarity 0.89)
- JetFormer: An autoregressive generative model of raw images and text (ICLR 2025 · similarity 0.88)
- Stabilize the Latent Space for Image Autoregressive Modeling: A Unified Perspective (NeurIPS 2024 · 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.