Enhance Photos With Deep Learning Part 2: The Data

In this article, we are going to look at some methods to degrade our data so it would be representative of production-time images. You can find the notebook for this article here. Without further ado, let's get coding! Finding data for a particular task is not easy: Not only must we come up with lots of input data, but we also need to come up with carefully-crafted labels to train the model. For instance, image segmentation requires tediously examining every pixel in a picture and identifying what object they belong to.