Skip to content

[unofficial] PyTorch Implementation of image quality assessment methods: IQA-CNN++ in ICIP2015 and IQA-CNN in CVPR2014

Notifications You must be signed in to change notification settings

lidq92/CNNIQAplusplus

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CNNIQAplusplus

PyTorch 1.3 implementation of the following paper: Kang, Le, et al. "Simultaneous estimation of image quality and distortion via multi-task convolutional neural networks." IEEE International Conference on Image Processing IEEE, 2015:2791-2795.

Note

The optimizer is chosen as Adam here, instead of the SGD with momentum in the paper.

Training

CUDA_VISIBLE_DEVICES=0 python main.py --exp_id=0 --database=LIVE --model=CNNIQAplusplus

Before training, the im_dir in config.yaml must to be specified.

Visualization

tensorboard --logdir=tensorboard_logs --port=6006 # in the server (host:port)
ssh -p port -L 6006:localhost:6006 user@host # in your PC. See the visualization in your PC

Requirements

conda create -n reproducibleresearch pip python=3.6
source activate reproducibleresearch
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
source deactive

Note: You need to install the right CUDA version.

TODO (If I have free time)

  • Simplify the code
  • Report results on some common databases
  • etc.