Research map / Image generation / Autoregressive and latent generation
PixelCNN Models with Auxiliary Variables for Natural Image Modeling
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 |
|---|---|---|---|
| Add a 4-bit quantized grayscale image as an auxiliary variable | PixelCNN · CIFAR-10 | visual quality of generated samples | Helped |
| Add a multi-resolution image-pyramid decomposition | PixelCNN · CelebA image sampling | sampling speed: 0.05 -> 0.004 seconds per pixel | Helped |
| Reduce the predicted mixture-component log-variance (ablation) | Pyramid PixelCNN · CelebA generated faces | perceptual quality | Helped |
| Replace the second random draw from a logistic component with its mode, while sampling the component by its mixture weight (ablation) | Pyramid PixelCNN · CelebA generated faces | perceptual quality | Helped |
| Model low- and high-level image details in separate distributions rather than sharing one model | PixelCNN · CIFAR-10 | global coherence of generated samples | Helped |
| Apply the conditional color model to real images’ quantized grayscale views (ablation) | Grayscale PixelCNN conditional model · CIFAR-10 test | visual plausibility of colorized images | No clear effect |
Closest papers by meaning
- PixelVAE: A Latent Variable Model for Natural Images (ICLR 2017 · similarity 0.89)
- PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modifications (ICLR 2017 · similarity 0.88)
- Pixel Recurrent Neural Networks (ICML 2016 · similarity 0.88)
- PixelNN: Example-based Image Synthesis (ICLR 2018 · similarity 0.88)
- Conditional Image Generation with PixelCNN Decoders (NeurIPS 2016 · similarity 0.87)
- Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction (NeurIPS 2024 · 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.