A dataset containing 84 images with no reference is provided along with their subjective scores. Their objective scores are to be calculated, and then PLCC, SRCC, KRCC values are to be calculated. The current best PLCC value is around 0.78 and beating this is a challenging part of this project. Any image processing techniques can be used but using machine learning / networks is not recommended due to lesser no of images in the dataset.
- opencv
- pandas
- numpy
- glob
- matplotlib
python3 main.py