This lecture explores advanced techniques for interactive and photorealistic image synthesis. After a compact repetition of rasterization fundamentals, we discuss central rendering effects such as reflections, transparency, and shadows, and then move on to specialized techniques for terrain rendering, fur, and the REYES pipeline.
The second part of the course connects classical image synthesis with modern neural approaches. We introduce neural functions and optimization, differentiable rendering, neural radiance fields, acceleration strategies, and 3D Gaussian splatting. The lecture concludes with neural appearance models and inverse rendering, where images are not only generated, but also used to recover and understand scene properties.
- verantwortliche Lehrperson: Jens Krüger
ePortfolio: Nein