3D Scene Reconstruction for Tele-robotics

Project Description

Scene reconstruction provides visual information of the remote site to the operator. The student will use visual-depth cameras and Nerf AI models to experiment the quality of the scene reconstruction with respect to camera resolution, training speed, and computing requirements. The expected outcome of the project will help analyze the fidelity of data quality and realtime-ness towards real-time tele-operation for scientific tasks. The student will participate in-person meetings with Argonne scientists and other in-person activities including seminars and poster presentations.

Project Objectives

• Topics: Scene Reconstruction Visual-Depth AI AR
• Skills: Python, PyTorch, Linux
• Difficulty: Moderate
• Size: Large (350 hours)
• Mentors: gemblerz seonghapark