Eyyub / Tensorflow Cyclegan
Lightweight CycleGAN tensorflow implementation π¦ <-> π
Stars: β 35
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Tensorflow Cyclegan
Image-Colorization-CycleGAN
Colorization of grayscale images using CycleGAN in TensorFlow.
Stars: β 16 (-54.29%)
Mutual labels: cyclegan
Attentiongan
AttentionGAN for Unpaired Image-to-Image Translation & Multi-Domain Image-to-Image Translation
Stars: β 341 (+874.29%)
Mutual labels: cyclegan
Cyclegan Tensorflow
Tensorflow implementation for learning an image-to-image translation without input-output pairs. https://arxiv.org/pdf/1703.10593.pdf
Stars: β 676 (+1831.43%)
Mutual labels: cyclegan
Pix2depth
DEPRECATED: Depth Map Estimation from Monocular Images
Stars: β 293 (+737.14%)
Mutual labels: cyclegan
Pycadl
Python package with source code from the course "Creative Applications of Deep Learning w/ TensorFlow"
Stars: β 356 (+917.14%)
Mutual labels: cyclegan
BicycleGAN-pytorch
Pytorch implementation of BicycleGAN with implementation details
Stars: β 99 (+182.86%)
Mutual labels: cyclegan
Contrastive Unpaired Translation
Contrastive unpaired image-to-image translation, faster and lighter training than cyclegan (ECCV 2020, in PyTorch)
Stars: β 822 (+2248.57%)
Mutual labels: cyclegan
Pytorch Cyclegan
A clean and readable Pytorch implementation of CycleGAN
Stars: β 558 (+1494.29%)
Mutual labels: cyclegan
traiNNer
traiNNer: Deep learning framework for image and video super-resolution, restoration and image-to-image translation, for training and testing.
Stars: β 130 (+271.43%)
Mutual labels: cyclegan
Generative models tutorial with demo
Generative Models Tutorial with Demo: Bayesian Classifier Sampling, Variational Auto Encoder (VAE), Generative Adversial Networks (GANs), Popular GANs Architectures, Auto-Regressive Models, Important Generative Model Papers, Courses, etc..
Stars: β 276 (+688.57%)
Mutual labels: cyclegan
Voice Converter Cyclegan
Voice Converter Using CycleGAN and Non-Parallel Data
Stars: β 384 (+997.14%)
Mutual labels: cyclegan
Pytorch-Image-Translation-GANs
Pytorch implementations of most popular image-translation GANs, including Pixel2Pixel, CycleGAN and StarGAN.
Stars: β 106 (+202.86%)
Mutual labels: cyclegan
Gans
Generative Adversarial Networks implemented in PyTorch and Tensorflow
Stars: β 714 (+1940%)
Mutual labels: cyclegan
automatic-manga-colorization
Use keras.js and cyclegan-keras to colorize manga automatically. All computation in browser. Demo is online:
Stars: β 20 (-42.86%)
Mutual labels: cyclegan
Gan Compression
[CVPR 2020] GAN Compression: Efficient Architectures for Interactive Conditional GANs
Stars: β 800 (+2185.71%)
Mutual labels: cyclegan
Von
[NeurIPS 2018] Visual Object Networks: Image Generation with Disentangled 3D Representation.
Stars: β 497 (+1320%)
Mutual labels: cyclegan
tensorflow-cyclegan
A lightweight CycleGAN tensorflow implementation.
If you plan to use a CycleGAN model for real-world purposes, you should use the Torch CycleGAN implementation.
Some examples
More lion2leopard (each classes contain only 100 instances!)
horse2zebra
horse2zebra failure
zebra2horse
wtf
More zebra2horse
apple2orange
See more in readme_imgs/
Build horse2zebra
- Download
horse2zebra.zip
from CycleGAN datasets - Unzip it here
.
- Run
python build_dataset.py horse2zebra/trainA horse2zebra/trainB trainA trainB
- (make sure
dataset_trainA.npy
&dataset_trainB.npy
are created) - Then, run
python example.py
- (If you want to stop and restart your training later you can do:
python example.py restore <last_iter_number>
)
Requiremennts
- Python 3.5
- Tensorflow
- Matplotlib
- Pillow
- (Only tested on Windows so far)
Very useful info
- Training took me ~1day (GTX 1060 3g)
- Each 100 steps the script adds an image in the
images/
folder - Each 1000 steps the model is saved in
models
- CycleGAN seems to be init-sensitive, if the generators only inverse colors: kill & re-try training
Todo
- [x] Image Pool
- [ ] Add learning reate linear decay
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].