Research map / Probabilistic and scientific modeling / Variational and generative sampling
Split Hierarchical Variational Compression
Tom Ryder, Chen Zhang, Ning Kang, Shifeng Zhang
Check 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 SHVC instead of other considered codecs | Lossless image compression · CLIC.mobile | BPD: 1.96$^{*}$ vs {2.26}$^*$ | Improved |
| Use SHVC instead of other considered codecs | Lossless image compression · CLIC.pro | BPD: 2.02$^{*}$ vs {2.44}$^*$ | Improved |
| Use SHVC instead of other considered codecs | Lossless image compression · DIV2K | BPD: 2.57$^{*}$ vs {2.57}$^*$ | No clear effect |
| Use SHVC instead of HiLLoC | CIFAR10 compression · CIFAR10 | inference time (s): 10.20 -> 4.63 | Improved |
| Replace the proposed downsampling operator with the usual space-to-depth transformation (ablation) | SHVC · CIFAR10 | BPD | Hurt |
| Increase autoregressive block size beyond k=4 (ablation) | SHVC · CIFAR10, ImageNet32, ImageNet64 | BPD | Hurt |
| Use autoregressive initial bits (ArIB) rather than ordinary SHVC initial bits (ablation) | SHVC · CIFAR10, ImageNet32, ImageNet64 single-image compression | additional bits per image: by close to 20x | Improved |
| Use stochastic rather than deterministic posterior sampling | VAE lossless compression models · CIFAR10, ImageNet32, ImageNet64 | BPD: ~5 % | Improved |
| Use repeated downsampling | VAE lossless compression models · CIFAR10, ImageNet32, ImageNet64 | BPD | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- iFlow — iFlow: Numerically Invertible Flows for Efficient Lossless Compression via a Uniform Coder (NeurIPS 2021) · better on 2 of 3 shared results
- Bit-Swap — Bit-Swap: Recursive Bits-Back Coding for Lossless Compression with Hierarchical Latent Variables (ICML 2019) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Joint Autoregressive and Hierarchical Priors for Learned Image Compression (NeurIPS 2018 · similarity 0.90)
- Variational image compression with a scale hyperprior (ICLR 2018 · similarity 0.89)
- Learned Lossless Image Compression based on Bit Plane Slicing (CVPR 2024 · similarity 0.87)
- Learned Image Transmission with Hierarchical Variational Autoencoder (AAAI 2025 · similarity 0.86)
- Improved Video VAE for Latent Video Diffusion Model (CVPR 2025 · similarity 0.86)
- SSC-VAE: Structured Sparse Coding Based Variational Autoencoder for Detail Preserved Image Reconstruction (AAAI 2025 · similarity 0.86)
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.
