Skip to content

Prototype-Based Multisource Domain Adaptation (TNNLS2021).

Notifications You must be signed in to change notification settings

maouestc/-Prototype-Based-Multisource-Domain-Adaptation

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

da

PyTorch implementation for Prototype-Based Multisource Domain Adaptation (TNNLS2021). This repository is based on framework from dassl and modified part of the code.

The installation can refer to dassl.

Please download dig-5, office-31, office-home to datasets folder.

Training:

CUDA_VISIBLE_DEVICES=1 python tools/train.py --trainer MSDTR_CDAN --source-domains mnist mnist_m svhn syn --target-domains usps  --dataset-config-file configs/datasets/digit5.yaml --config-file configs/trainers/msdtr_cdan/digit5.yaml --output-dir output_final/msdtr_cdan/dig/usps/1

Citation

If you use this code for your research, please cite our paper:

@article{zhou2021prototype,
  title={Prototype-Based Multisource Domain Adaptation},
  author={Zhou, Lihua and Ye, Mao and Zhang, Dan and Zhu, Ce and Ji, Luping},
  journal={IEEE Transactions on Neural Networks and Learning Systems},
  year={2021},
  publisher={IEEE}
}

About

Prototype-Based Multisource Domain Adaptation (TNNLS2021).

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%