GLOW is a type of flow-based generative model that is based on an invertible $1 \times 1$ convolution. This builds on the flows introduced by NICE and RealNVP. It consists of a series of steps of flow, combined in a multi-scale architecture; see the Figure to the right. Each step of flow consists of Act Normalization followed by an invertible $1 \times 1$ convolution followed by an affine coupling layer.
Source: Glow: Generative Flow with Invertible 1x1 ConvolutionsPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Image Dehazing | 2 | 11.76% |
Image Enhancement | 2 | 11.76% |
Pseudo Label | 1 | 5.88% |
Offline RL | 1 | 5.88% |
Benchmarking | 1 | 5.88% |
Molecular Docking | 1 | 5.88% |
Pose Prediction | 1 | 5.88% |
Low-Light Image Enhancement | 1 | 5.88% |
Zero-Shot Learning | 1 | 5.88% |
Component | Type |
|
---|---|---|
Activation Normalization
|
Normalization | |
Affine Coupling
|
Bijective Transformation | |
Invertible 1x1 Convolution
|
Convolutions | |
Normalizing Flows
|
Distribution Approximation |