Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Contrastive unpaired image-to-image translation, faster and lighter training than cyclegan (ECCV 2020, in PyTorch)
| Date | Stars |
|---|---|
| 2026-07-24 | 2537 |
| 2026-07-25 | 2538 |
| 2026-07-28 | 2539 |
| 2026-07-30 | 2539 |
| 2026-08-06 | 2539 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Contrastive Unpaired Translation (CUT)
### [video (1m)](https://youtu.be/Llg0vE_MVgk) | [video (10m)](https://youtu.be/jSGOzjmN8q0) | [website](http://taesung.me/ContrastiveUnpairedTranslation/) | [paper](https://arxiv.org/pdf/2007.15651)
<br>
<img src='imgs/gif_cut.gif' align="right" width=960>
<br><br><br>
We provide our PyTorch implementation of unpaired image-to-image translation based on patchwise contrastive learning and adversarial learning. No hand-crafted loss and inverse network is used. Compared to [CycleGAN](https://github.com/junyanz/CycleGAN), our model training is faster and less memory-intensive. In addition, our method can be extended to single image training, where each “domain” is only a *single* image.
[Contrastive Learning for Unpaired Image-to-Image Translation](http://taesung.me/ContrastiveUnpairedTranslation/)
[Taesung Park](https://taesung.me/), [Alexei A. Efros](https://people.eecs.berkeley.edu/~efros/), [Richard Zhang](https://richzhang.github.io/), [Jun-Yan Zhu](https://www.cs.cmu.edu/~junyanz/)<br>
UC Berkeley and Adobe Research<br>
In ECCV 2020
<img src='imgs/patchnce.gif' align="right" width=960>
<br><br><br>
### Pseudo code
```python
import torch
cross_entropy_loss = torch.nn.CrossEntropyLoss()
# Input: f_q (BxCxS) and sampled features from H(G_enc(x))
# Input: f_k (BxCxS) are sampled features from H(G_enc(G(x))
# Input: tau is the temperature used in PatchNCE loss.
# Output: PatchNCE loss
def PatchNCELoss(f_q, f_k, tau=0.07):
# batch size, channel size, and number of sample locations
B, C, S = f_q.shape
# calculate v * v+: BxSx1
l_pos = (f_k * f_q).sum(dim=1)[:, :, None]
# calculate v * v-: BxSxS
l_neg = torch.bmm(f_q.transpose(1, 2), f_k)
# The diagonal entries are not negatives. Remove them.
identity_matrix = torch.eye(S)[None, :, :]
l_neg.masked_fill_(identity_matrix, -float('inf'))
# calculate logits: (B)x(S)x(S+1)
logits = torch.cat((l_pos, l_neg), dim=2) / tau
# return PatchNCE loss
predictions = logits.flatten(0, 1)
targets = torch.zeros(B * S, dtype=torch.long)
return cross_entropy_loss(predictions, targets)
```
## Example Results
### Unpaired Image-to-Image Translation
<img src="imgs/results.gif" width="800px"/>
### Single Image Unpaired Translation
<img src="imgs/singleimage.gif" width="800px"/>
### Russian Blue Cat to Grumpy Cat
<img src="imgs/grumpycat.jpg" width="800px"/>
### Parisian Street to Burano's painted houses
<img src="imgs/paris.jpg" width="800px"/>
## Prerequisites
- Linux or macOS
- Python 3
- CPU or NVIDIA GPU + CUDA CuDNN
### Update log
9/12/2020: Added single-image translation.
### Getting started
- Clone this repo:
```bash
git clone https://github.com/taesungp/contrastive-unpaired-translation CUT
cd CUT
```
- Install PyTorch 1.1 and other dependencies (e.g., torchvision, visdom, dominate, gputil).
For pip users, please type the command `pip install -r requirements.txt`.
For Conda users, you can create a new Conda environment using `conda env create -f environment.yml`.
### CUT and FastCUT Training and Test
- Download the `grumpifycat` dataset (Fig 8 of the paper. Russian Blue -> Grumpy Cats)
```bash
bash ./datasets/download_cut_dataset.sh grumpifycat
```
The dataset is downloaded and unzipped at `./datasets/grumpifycat/`.
- To view training results and loss plots, run `python -m visdom.server` and click the URL http://localhost:8097.
- Train the CUT model:
```bash
python train.py --dataroot ./datasets/grumpifycat --name grumpycat_CUT --CUT_mode CUT
```
Or train the FastCUT model
```bash
python train.py --dataroot ./datasets/grumpifycat --name grumpycat_FastCUT --CUT_mode FastCUT
```
The checkpoints will be stored at `./checkpoints/grumpycat_*/web`.
- Test the CUT model:
```bash
python test.py --dataroot ./datasets/grumpifycat --name grumpycat_CUT --CUT_mode CUT --phase train
```
The test results will be saved to a html file here: `./results/grumpiExcerpt of 14,072 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:f748b64fe703c210, topic:pytorch
matched fp:f748b64fe703c210, topic:computer-vision
matched fp:f748b64fe703c210, topic:image-generation
