Neural Scene Representations for Learning-based View Synthesis and Its Applications

Explore a seminar on neural scene representations for learning-based view synthesis and its applications. Delve into the speaker's background in 3D computer vision, computer graphics, and deep learning. Discover how neural scene representations offer a more efficient alternative to traditional 3D representations for creating lifelike images and videos. Learn about the advantages of neural scene representations, including their ability to be trained end-to-end from data and tailored to specific tasks such as novel view synthesis, relighting, scene composition, text-to-3D, and text-to-4D. Gain insights into the latest advancements in computer graphics and their potential applications in mixed reality.