Created on 9th May 2020
•
In Machine and Deep learning, when working with images the quality of dataset matters a lot. When one downloads images from google using scripts, image becomes blur, thus reducing datasets quality. So, we have created a model which will minimize blur effect in images for better usage.
Choosing the activation function was a tricky task as it was not a classification problem.
Optimum number of network layers was another.
GPU access and training takes a lot of time so have to be accurate for best performance
Memory availability
Technologies used