Skip to content

Official inference code of CVPR2021 paper "Deep Learning-based Distortion Sensitivity Prediction for Full-Reference Image Quality Assessment"

Notifications You must be signed in to change notification settings

fdp0525/DeepQA-modified

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FR-IQA condisdering distortion sensitivity

This is the test code for NTIRE Perceptual Image Quality Assessment (PIQA) Challenge

By running "test_FR.py", result of inference is saved on "output.txt".

Before running "test_FR.py" you have to consider two things.

  1. There are four parameters in "test_FR.py" (from line 132 to 146)

    1. GPU_NUM: name of GPU you want to use.
      • ex) GPU_NUM = "0" or GPU_NUM="2"
    2. dirname: forder directory where test image exist
    3. weights_file: file name of model weights
    4. result_score_txt: text file name for storing inference results

    if you setting 1), 2), 3) and running the "test_FR.py", inference result is saved in 4)

  2. Test_Images folder You must save the reference and distoted images in "Test_Images" folder The example of saving file is listed as below

    1. put reference images in "Test_Images/Reference" folder
    2. put distorted images in "Test_Images folder"

    There are examplar images in "Test_Images" foler

About

Official inference code of CVPR2021 paper "Deep Learning-based Distortion Sensitivity Prediction for Full-Reference Image Quality Assessment"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%