NL-CSNet-Pytorch
Pytorch code for paper "Image Compressed Sensing Using Non-local Neural Network" TMM 2021.
Note: this repo only shows the strategy of plugging the Non-local module (with non-local coupling loss constraint) into a simple CNN-based CS network (in the measurement domain and feature domain). For more details of the NL-CSNet atchitecture, please refer to the paper.
Framework
Requirements and Dependencies
- Ubuntu 16.04 CUDA 10.0
- Python3 （Testing in Python3.5）
- Pytorch 1.1.0
- Torchvision 0.2.2
How to Run
Training NL-CSNet
-
Preparing the dataset for training
-
Editing the path of training data in file
train.py. -
For NL-CSNet training in terms of subrate=0.1:
python train.py --sub_rate=0.1 --block_size=32
Testing NL-CSNet
-
Preparing the dataset for testing
-
Editing the path of trained model in file
test.py. -
For NL-CSNet testing in terms of subrate=0.1:
python test.py --cuda --sub_rate=0.1 --block_size=32
NL-CSNet results
Subjective results
Objective results
Additional instructions
- For training data, you can choose any natural image dataset.
- If you like this repo, Star or Fork to support my work. Thank you.
- If you have any problem for this code, please email: [email protected]
84 Jan 4, 2023
22 Dec 12, 2022
28 Aug 29, 2022
120 Dec 12, 2022
1.5k Dec 28, 2022
148 Dec 30, 2022
![[TOG 2021] PyTorch implementation for the paper: SofGAN: A Portrait Image Generator with Dynamic Styling.](https://github.com/apchenstu/apchenstu.github.io/raw/master/sofgan/img/semantic_level.png)
694 Dec 23, 2022
258 Jan 2, 2023
23 Nov 5, 2022
3 Jan 3, 2023
366 Dec 28, 2022
198 Dec 27, 2022
25 Sep 6, 2022
18 Oct 6, 2022
585 Jan 4, 2023
111 Dec 31, 2022
100 Dec 1, 2022
574 Jan 2, 2023