Research map / Image generation / Autoregressive and latent generation
Bit Prioritization in Variational Autoencoders via Progressive Coding
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 |
|---|---|---|---|
| two-stage progressive coding training (distortion-targeted first stage, then conditional second stage) | HVAE (VDVAE-based) · SVHN | FID: 18.92 -> 10.07 | Helped |
| two-stage progressive coding training | HVAE (VDVAE-based) · CelebA | FID: 13.33 -> 8.54 | Helped |
| two-stage progressive coding training | HVAE (VDVAE-based) · LSUN-Bedrooms | FID: 40.71 -> 36.87 | Helped |
| two-stage progressive coding training | HVAE (VDVAE-based) · SVHN | ELBO (per-dim): -1.31 -> -1.46 | Hurt |
| two-stage progressive coding training | HVAE (VDVAE-based) · CelebA | ELBO (per-dim): -1.44 -> -1.58 | Hurt |
| two-stage progressive coding training | HVAE (VDVAE-based) · LSUN-Bedrooms | ELBO (per-dim): -1.72 -> -1.78 | Hurt |
| choosing rate target RK=0 for the first K latents (distortion-targeting at extreme) | HVAE progressive coding · text claim | FID | Hurt |
Closest papers by meaning
- Hierarchical Quantized Autoencoders (NeurIPS 2020 · similarity 0.91)
- Trading Information between Latents in Hierarchical Variational Autoencoders (ICLR 2023 · similarity 0.90)
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.89)
- Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders (ICLR 2024 · similarity 0.88)
- Be a Goldfish: Forgetting Bad Conditioning in Sparse Linear Regression via Variational Autoencoders (ICML 2025 · similarity 0.88)
- Learned Image Transmission with Hierarchical Variational Autoencoder (AAAI 2025 · similarity 0.88)
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.