Image-Colorization with GANs
Faizaan Al Faisal
Published
0
Deep Learning Model for Image Restoration and Enhancement
Built a Generative Adversarial Network (GAN) with PyTorch to colorize grayscale images using the COCO dataset. Implemented a UNet-based generator and a PatchGAN discriminator, trained on Kaggle for image-to-image translation. Demonstrated the ability to reconstruct mostly realistic colors in natural scenes.