PixelCNN

Introduced by Oord et al. in Conditional Image Generation with PixelCNN Decoders

A PixelCNN is a generative model that uses autoregressive connections to model images pixel by pixel, decomposing the joint image distribution as a product of conditionals. PixelCNNs are much faster to train than PixelRNNs because convolutions are inherently easier to parallelize; given the vast number of pixels present in large image datasets this is an important advantage.

Source: Conditional Image Generation with PixelCNN Decoders

Latest Papers

PAPER DATE
Nonlinear Equation Solving: A Faster Alternative to Feedforward Computation
Yang SongChenlin MengRenjie LiaoStefano Ermon
2020-02-10
Multimodal Controller for Generative Models
Enmao DiaoJie DingVahid Tarokh
2020-02-07
Closing the Dequantization Gap: PixelCNN as a Single-Layer Flow
| Didrik NielsenOle Winther
2020-02-06
Hilbert-Based Generative Defense for Adversarial Examples
Yang Bai Yan Feng Yisen Wang Tao Dai Shu-Tao Xia Yong Jiang
2019-10-01
PixelVAE++: Improved PixelVAE with Discrete Prior
Hossein SadeghiEvgeny AndriyashWalter VinciLorenzo BuffoniMohammad H. Amin
2019-08-26
Bayesian Volumetric Autoregressive generative models for better semisupervised learning
| Guilherme PomboRobert GrayTom VarsavskyJohn AshburnerParashkev Nachev
2019-07-26
Generating Diverse High-Fidelity Images with VQ-VAE-2
| Ali RazaviAaron van den OordOriol Vinyals
2019-06-02
ShieldNets: Defending Against Adversarial Attacks Using Probabilistic Adversarial Robustness
Rajkumar Theagarajan Ming Chen Bir Bhanu Jing Zhang
2019-06-01
Tree Tensor Networks for Generative Modeling
Song ChengLei WangTao XiangPan Zhang
2019-01-08
Practical Full Resolution Learned Lossless Image Compression
| Fabian MentzerEirikur AgustssonMichael TschannenRadu TimofteLuc Van Gool
2018-11-30
The Variational Homoencoder: Learning to learn high capacity generative models from few examples
| Luke B. HewittMaxwell I. NyeAndreea GaneTommi JaakkolaJoshua B. Tenenbaum
2018-07-24
Representation Learning with Contrastive Predictive Coding
| Aaron van den OordYazhe LiOriol Vinyals
2018-07-10
Feature Map Variational Auto-Encoders
Lars MaaløeOle Winther
2018-01-01
The Variational Homoencoder: Learning to Infer High-Capacity Generative Models from Few Examples
Luke HewittAndrea GaneTommi JaakkolaJoshua B. Tenenbaum
2018-01-01
Spatial PixelCNN: Generating Images from Patches
Nader AkouryAnh Nguyen
2017-12-03
Few-shot Autoregressive Density Estimation: Towards Learning to Learn Distributions
Scott ReedYutian ChenThomas PaineAäron van den OordS. M. Ali EslamiDanilo RezendeOriol VinyalsNando de Freitas
2017-10-27
PixColor: Pixel Recursive Colorization
Sergio GuadarramaRyan DahlDavid BieberMohammad NorouziJonathon ShlensKevin Murphy
2017-05-19
Parallel Multiscale Autoregressive Density Estimation
Scott ReedAäron van den OordNal KalchbrennerSergio Gómez ColmenarejoZiyu WangDan BelovNando de Freitas
2017-03-10
Count-Based Exploration with Neural Density Models
Georg OstrovskiMarc G. BellemareAaron van den OordRemi Munos
2017-03-03
PixelCNN Models with Auxiliary Variables for Natural Image Modeling
Alexander KolesnikovChristoph H. Lampert
2016-12-24
PixelVAE: A Latent Variable Model for Natural Images
Ishaan GulrajaniKundan KumarFaruk AhmedAdrien Ali TaigaFrancesco VisinDavid VazquezAaron Courville
2016-11-15
Conditional Image Generation with PixelCNN Decoders
| Aaron van den OordNal KalchbrennerOriol VinyalsLasse EspeholtAlex GravesKoray Kavukcuoglu
2016-06-16

Components

COMPONENT TYPE
🤖 No Components Found You can add them if they exist; e.g. Mask R-CNN uses RoIAlign

Categories