Skip to content

Latest commit

 

History

History
148 lines (99 loc) · 5.74 KB

README.md

File metadata and controls

148 lines (99 loc) · 5.74 KB
Raw

C-SFDA: A Curriculum Learning Aided Self-Training Framework for Efficient Source Free Domain Adaptation

If you like our project, please give us a star ⭐ on GitHub for the latest update.

arXiv License: MIT

Paper

First, Create an Environment

conda create -n domain_ada 
conda activate domain_ada

pip3 install torch==1.8.1+cu111 torchvision==0.9.1+cu111 -f https://download.pytorch.org/whl/torch_stable.html
pip install hydra-core numpy omegaconf sklearn tqdm wandb

For VisDA-C

Prepare dataset

Please download the VisDA-C dataset, and put it under ${DATA_ROOT}. For your convenience we also compiled .txt files based on the the image labels, provided under ./data/VISDA-C/. The prepared directory would look like:

${DATA_ROOT}
├── VISDA-C
│   ├── train
│   ├── validation
│   ├── train_list.txt
│   ├── validation_list.txt

${DATA_ROOT} is set to ./data/ by default, which can be modified in configs/data/basic.yaml or via hydra command line interface data.data_root=${DATA_ROOT}.

Training

We use hydra as the configuration system. By default, the working directory is ./output, which can be changed directly from configs/root.yaml or via hydra command line interface workdir=${WORK_DIR}.

VISDA-C experiments are done for train to validation adaptation. Before the adaptation, we should have the source model. You may train the source model with script train_source_VisDA.sh as shown below.

export CUDA_VISIBLE_DEVICES=0,1,2,3
bash train_source_VisDA.sh

We also provide the link to pre-trained source models from 3 seeds (2020, 2021, 2022) which can be downloaded from here.

After obtaining the source models, put them under ${SRC_MODEL_DIR} and run

export CUDA_VISIBLE_DEVICES=0,1,2,3
bash train_target_VisDA.sh

For DomainNet-126

Prepare dataset

Please download the DomainNet dataset (cleaned version), and put it under ${DATA_ROOT}. Notice that we follow MME to use a subset that contains 126 classes from 4 domains, so we also compiled .txt files for your convenience based on the the image labels, provided under ./data/domainnet-126/. The prepared directory would look like:

${DATA_ROOT}
├── domainnet-126
│   ├── real
│   ├── sketch
│   ├── clipart
│   ├── painting
│   ├── real_list.txt
│   ├── sketch_list.txt
│   ├── clipart_list.txt
│   ├── painting_list.txt

${DATA_ROOT} is set to ./data/ by default, which can be modified in configs/data/basic.yaml or via hydra command line interface data.data_root=${DATA_ROOT}.

Training

We use hydra as the configuration system. By default, the working directory is ./output, which can be changed directly from configs/root.yaml or via hydra command line interface workdir=${WORK_DIR}.

DomainNet-126 experiments are done for 7 domain shifts constructed from combinations of Real, Sketch, Clipart, and Painting. Before the adaptation, we should have the source model. You may train the source model with script train_source_DomainNet.sh as shown below.

export CUDA_VISIBLE_DEVICES=0,1,2,3
bash train_source_DomainNet.sh <SOURCE_DOMAIN>

We also provide the link to pre-trained source models from 3 seeds (2020, 2021, 2022) which can be downloaded from here.

After obtaining the source models, put them under ${SRC_MODEL_DIR} and run

export CUDA_VISIBLE_DEVICES=0,1,2,3
bash train_target_DomainNet.sh <SOURCE_DOMAIN> <TARGET_DOMAIN> <SRC_MODEL_DIR>

For Office-Home and Office-31

  • We follow SHOT Paper GitHub Repo for these two datasets

  • Check the "Office" Folder

cd Office

  • Please manually download the datasets Office, Office-Home

  • To train the source models

export CUDA_VISIBLE_DEVICES=0,1,2,3
bash office-home_source.sh
  • For Adaptation
export CUDA_VISIBLE_DEVICES=0,1,2,3
bash office_home_target.sh

Note

  • If the simulation ends without any error, set HYDRA_FULL_ERROR=1 GPU Usage: 2 NVIDIA RTX A40 GPUs

  • This repo is built on AdaContrast and SHOT GitHub repos

Reference

If you find this work helpful to your own work, please consider citing us:

@inproceedings{karim2023c,
  title={C-sfda: A curriculum learning aided self-training framework for efficient source free domain adaptation},
  author={Karim, Nazmul and Mithun, Niluthpol Chowdhury and Rajvanshi, Abhinav and Chiu, Han-pang and Samarasekera, Supun and Rahnavard, Nazanin},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={24120--24131},
  year={2023}
}