Image Generation

Recent generative models are considered transformative tools that are redefining the field of artificial intelligence and computing in general. These models, often based on deep neural networks, play a central role in tasks such as image generation, rendering, and even the creation of creative content. Their significance lies in their ability to learn complex patterns and relationships from vast datasets, enabling the generation of realistic and diverse results. This course focuses on the latest image generation techniques, which involve rendering processes using neural networks. Neural rendering is a key aspect of this course and can be roughly defined as a technique in which traditional rendering pipelines (such as ray tracing, volumetric rendering, image-based rendering, etc.) are augmented with neural networks. We will first review the fundamentals of computer graphics and deep neural networks, focusing on 2D image-based neural rendering. We will then discuss advances in 3D volumetric rendering, including Neural Radiance Fields (NeRFs).