Skip to content

Pytorch codes for NeurIPS 2021 paper titled "Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data".

Notifications You must be signed in to change notification settings

asrafulashiq/dynamic-cdfsl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data

Pytorch implementation of NeurIPS 2021 paper Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data.

Paper Poster slides

Proposed Framework

Prerequisite

Install the required python packages by:

pip install -r requirements.txt

Dataset

Download the datasets from here, and keep them in the data/ directory.

Pretraining

Download the cross-entropy pretrained model (download from here) on mini-ImageNet in ckpt/ce_miniImageNet_resnet10.

Or, train by running:

python main.py system=ce  backbone=resnet10 data.dataset=miniImageNet_train  model_name=ce_miniImageNet_resnet10 trainer.gpus=4

Training

To train on mini-ImageNet and unlabeled target images, run the following command:

python main.py system=ce_distill_ema_sgd trainer.gpus=1 backbone=resnet10 \
  data.val_dataset=EuroSAT_test data.test_dataset=null print_val=false \
  trainer.log_every_n_steps=-1 \
  unlabel_params.dataset=EuroSAT_train data.num_episodes=600 \
  trainer.progress_bar_refresh_rate=0 print_val=false launcher.gpus=1\
  model_name=dynamic_cdfsl_EuroSAT

Change EuroSAT to other dataset to train on CropDisese, ChestX, ISIC.

Model will be saved in ckpt/dynamic_cdfsl_EuroSAT/last.ckpt

Pretrained checkpoints

You can download the pretrained checkpoints on the BSCD-FSL benchmark datasets from here.

Few-shot evaluation

For 5-way 5-shot evaluation on EuroSAT dataset:

python main.py system=few_shot  data.test_dataset=EuroSAT_test  ckpt=[pretrained-checkpoint]

Citation

If you find this repo useful for your research, please consider citing the paper:

@inproceedings{
islam2021dynamic,
title={Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data},
author={Ashraful Islam and Chun-Fu Chen and Rameswar Panda and Leonid Karlinsky and Rogerio Feris and Richard Radke},
booktitle={Thirty-Fifth Conference on Neural Information Processing Systems},
year={2021},
url={https://openreview.net/forum?id=_4VxORHq-0g}
}

Acknowledgement

About

Pytorch codes for NeurIPS 2021 paper titled "Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data".

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published