Scalable and Controllable Creation of Digital Twins for Embodied AI

Speaker:              Manolis Savva

Time:                   14:00, Oct. 26th.

Location:             SIST 1A-200

Host:                    Prof. Kai Wang

Abstract:

Simulated indoor 3D scenes are increasingly used to develop and evaluate embodied AI agents and autonomous robot systems. However, scalable, physically-grounded, and controllable creation of digital twins for real-world objects and scenes that is necessary for such simulations is a key bottleneck. In this talk, I will describe recent projects that aim to alleviate this bottleneck, focusing particularly on physically-grounded interactive 3D objects, and controllable generation of 3D scenes.

Bio:

Manolis Savva is an Associate Professor at Simon Fraser University, and a Canada Research Chair in Computer Graphics.  His research focuses on analysis, organization and generation of 3D content.  Prior to his current position he was a visiting researcher at Facebook AI Research, and a postdoctoral researcher at Princeton University.  He received his Ph.D. from Stanford University under the supervision of Pat Hanrahan.  His work has been recognized through several awards including an ACM UIST notable paper award (ReVision), an ICCV best paper nomination (Habitat), the 2022 Graphics Interface early career researcher award, an ICLR 2023 outstanding paper award (Emergence of Maps), and a CS-CAN 2025 Early Career Research Award.