Research map / Image generation / Image restoration
Variable Rate Deep Image Compression With a Conditional Autoencoder
Yoojin Choi, Mostafa El-Khamy, Jungwon Lee
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 |
|---|---|---|---|
| train with universal quantization instead of additive uniform noise (ablation) | variable-rate conditional autoencoder · Kodak (24 images) | PSNR | Improved |
| mix quantization bin sizes in [0.5, 2] instead of a wider range (ablation) | variable-rate conditional autoencoder · Kodak (24 images), lambda=10^-2.5 | PSNR | Improved |
| widen the range of mixed bin sizes (ablation) | variable-rate conditional autoencoder · Kodak (24 images), lambda=10^-2.5 | PSNR | Hurt |
| use one variable-rate network instead of separate fixed-rate networks per lambda and bin size | variable-rate conditional autoencoder · Kodak (24 images) | PSNR and MS-SSIM | No clear effect |
| use the variable-rate model optimized for MSE instead of BPG | variable-rate conditional autoencoder · Kodak (24 images) | PSNR and MS-SSIM: 0.3 dB PSNR gain | Improved |
| make lambda a continuous conditioning input instead of one-hot over a finite set (ablation) | variable-rate conditional autoencoder · Kodak (24 images) | R-D performance | Hurt |
Closest papers by meaning
- Conditional Probability Models for Deep Image Compression (CVPR 2018 · similarity 0.89)
- Flexible Neural Image Compression via Code Editing (NeurIPS 2022 · similarity 0.88)
- End-to-end Optimized Image Compression (ICLR 2017 · similarity 0.88)
- Lossy Image Compression with Compressive Autoencoders (ICLR 2017 · similarity 0.88)
- Selective compression learning of latent representations for variable-rate image compression (NeurIPS 2022 · similarity 0.88)
- Improving Inference for Neural Image Compression (NeurIPS 2020 · 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.
