https://github.com/naoto0804/pytorch-domain-adaptation
Unofficial pytorch implementation of algorithms for domain adaptation
https://github.com/naoto0804/pytorch-domain-adaptation
cnn domain-adaptation generative-adversarial-network pytorch
Last synced: about 2 months ago
JSON representation
Unofficial pytorch implementation of algorithms for domain adaptation
- Host: GitHub
- URL: https://github.com/naoto0804/pytorch-domain-adaptation
- Owner: naoto0804
- Created: 2018-04-16T04:43:28.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2018-08-15T08:24:38.000Z (about 7 years ago)
- Last Synced: 2025-04-05T05:24:09.669Z (6 months ago)
- Topics: cnn, domain-adaptation, generative-adversarial-network, pytorch
- Language: Python
- Homepage:
- Size: 2.34 MB
- Stars: 27
- Watchers: 1
- Forks: 7
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# pytorch-domain-adaptation
This is an unofficial pytorch implementation of algorithms for domain adaptation.
**Note that this is an ongoing project and I cannot fully reproduce the results. Suggestions are welcome!**
## List of algorithms
- From source to target and back: symmetric bi-directional adaptive GAN [Russo+, CVPR2018].
- Augmented Cyclic Adversarial Learning for Domain Adaptation [Hosseini-Asl+, arXiv2018].
## Requirements
- Python 3.5+
- PyTorch 0.4
- TorchVision
- TensorboardX
- batchup
- click
## Usage
These examples are for the MNIST to USPS experiment.
### Train `Source Only` Model
```
CUDA_VISIBLE_DEVICES= python train_classifier.py --exp mnist_usps --train_type unsup
```
### Train `Target Only` Model
```
CUDA_VISIBLE_DEVICES= python train_classifier.py --exp mnist_usps --train_type sup
```
### Train Model
```
UDA_VISIBLE_DEVICES= python test_classifier.py --exp mnist_usps --snapshot
```