This project explores the use of Generative Adversarial Networks (GANs) for image colorization and compares the performance of a standard GAN with a Wasserstein-GAN (WGAN) structure. The experiments are conducted on the Tiny-ImageNet-200 dataset, offering insights into the challenges of using an uncommon dataset for this task. The study showcases the adaptability of GANs for image processing and provides valuable comparisons between WGAN and GAN models. It also highlights the exploration of the Tiny-ImageNet-200 dataset for image colorization.
This project was developed in February 2023 for the course Neural Networks and Deep Learning at the University of Padova (academic year: 2022-2023). Team members:
- Tonazzo Valentina: https://github.com/ValentinaTonazzo
- Zoppellari Elena: https://github.com/zoppellarielena
In the repository also the report about the project and a brief presentation of the project are provided as pdf file.