HiNAS

Code for paper:

Memory-Efficient Hierarchical Neural Architecture Search for Image Denoising (CVPR 2020)

Memory-Efficient Hierarchical Neural Architecture Search for Image Restoration

Compared with the mehtod propsoed in Memory-Efficient Hierarchical Neural Architecture Search for Image Denoising, we made some improvements, which are explained in "log file for DNAS_For_IR"

Requirements

This code is tested on Pytorch = 1.0.0.

I build my experimental environment by create a virtual env via anaconda3. 
After activating you env, you can install other all dependences via run: pip install -r requirements.txt. 
Note that,  to install graphviz, you also need to run: conda install graphviz. 

Searching

#searching for denoising network
cd ./tools/
python search.py --config-file "../configs/dn/BSD500_3c4n/03_search_CR_R0.yaml" --device '0'

#searching for super-resolution network
cd ./tools/
python search.py --config-file "../configs/sr/DIV2K_3c3n/03_search_CR.yaml" --device '0'

Training

#training the founded denosing network with noise factor=30
cd ./tools/
python train.py --config-file "../configs/dn/BSD500_3c4n/03_train_CR_RO/train_s30.yaml" --device '0'

#training the founded super-resolution network with SR factor=3
cd ./tools/
python train.py --config-file "../configs/sr/DIV2K_2c3n/03_x3_train_CR.yaml" --device '0'

Inference

# testing the trained denoising network with noise factor=[30 50 70]
cd ./tools/
python dn_eval.py --config-file "../configs/dn/BSD500_3c4n/03_train_CR_RO/03_infe.yaml" --device '0'

# testing the trained super-resolution network with sr factor=3
cd ./tools/
python sr_eval.py --config-file "../configs/sr/DIV2K_2c3n/03_x3_infe_CR.yaml" --device '0'

Datasets

Fetch code:1111

Denoising datasets: BSD200 BSD300

Super-resolution datasets: DIV2K_800 DIV2K_100 BSD100 BSDS100 Urban100 Manga109 General100 Set14 Set5

Citation

If you use this code in your paper, please cite our papers

@inproceedings{zhang2020memory,
  title={Memory-efficient hierarchical neural architecture search for image denoising},
  author={Zhang, Haokui and Li, Ying and Chen, Hao and Shen, Chunhua},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={3657--3666},
  year={2020}
}