Automatic Image Colorization using U-Net
Divyansh Goel, Sakshi Jain, Dinesh Kumar Vishwakarma, Aryan Bansal · 2021
Automatic Image Colorization is the process of converting a gray-scale image to a corresponding colored image as output, without any human interference. The main objective of the research is to develop an automated technique that colorizes a given gray-scale image and generates a colorized version of the image as the output. We have presented our custom U-Net architecture model. We have treated this problem as an image segmentation + multinomial classification problem. The results contain handpicked images from our outputs. We have included the black and white image, our model's colored output and the colored ground truth of the images.