Some models with easy understanding code,it will help you understand what the model does.These models can only work with not very good results.
1.Adversarial Networks (GAN)
- AAE Adversarial Autoencoders
- ACGAN Conditional Image Synthesis With Auxiliary Classifier GANs
- Auto encoder Recent Advances in Autoencoder-Based Representation Learning
- BGAN Boundary-Seeking Generative Adversarial Networks
- BiGAN Bidirectional Generative Adversarial Network
- CCGAN Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks
- CGAN Conditional Generative Adversarial Nets
- CoGAN Coupled generative adversarial networks
- CycleGAN Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
- DCGAN Deep Convolutional Generative Adversarial Network
- GAN Generative Adversarial Network with a MLP generator and discriminator
- VAE Auto-Encoding Variational Bayes
2.Special structure Convolutional network (Not recurring paper)
- DenseNet Densely Connected Convolutional Networks
- ResNet Deep Residual Learning for Image Recognition
- HighwayNet Highway Networks
- MobileNet_v1&v2 MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
MobileNetV2: Inverted Residuals and Linear Bottlenecks
3.Other Models
- None
The results are based on MNIST or Fashion-MNIST
Input image → Hidden layer → Output image
Real image → Random cropping image → Repaired image
A: MNIST B: Rotate 90 degrees MNIST
The model try to convert between A and B.
A → B
B → A
A: MNIST B: Rotate 90 degrees MNIST
The model try to convert between A and B.
A → B
B → A
Input image → Hidden layer → Output image
