About this document
Learning Functors Using Gradient Descent by celyjenrique is a document available to read on EtoBox.
This paper presents a category-theoretic framework for understanding CycleGAN, a neural network for unpaired image-to-image translation. It introduces the concept of schemas as categories defined by generators and relations, allowing for the learning of functors rather than just functions through gradient descent. The proposed system can learn to manipulate images, such as inserting and deleting objects, without the need for paired data, demonstrating promising results on the CelebA dataset.
- Author
- celyjenrique
- Language
- EN