Seminar: Pixel Club
Clicks and Prompts: Intuitive Interactive Segmentation and Local Texturing of 3D Shapes
Lecturer:
Itai Lang
Research Areas:
| 3D digital content is widely used in fields such as engineering, medicine, and entertainment, with a rapidly growing demand for 3D asset processing and manipulation. However, current tools remain largely inaccessible, requiring expert knowledge and extensive manual effort. Meanwhile, 2D foundation models have revolutionized image processing and generation, offering intuitive interfaces through point clicks and text prompts. This raises a fundamental question: How can we bridge the gap between these two worlds?
In this talk, I will demonstrate how to democratize 3D content understanding and editing by leveraging the powerful prior knowledge of 2D foundation models. I will discuss the challenges of applying these models to interactive segmentation and local editing of 3D shapes and introduce novel techniques we have developed to overcome these hurdles. Finally, I will showcase new creative opportunities in geometry editing and share my perspective on the future of 3D content manipulation. |
| Itai Lang (https://itailang.github.io/) is a postdoctoral researcher at the University of Chicago, working with Professor Rana Hanocka. He earned his PhD from Tel Aviv University under the guidance of Professor Shai Avidan. His research focuses on developing artificial intelligence for geometry processing, spanning both generative and discriminative tasks such as local texture editing, interactive segmentation, geometric stylization, and more. He is particularly passionate about creating innovative solutions to make 3D content manipulation more intuitive and accessible. |