Re-imagining Robot Autonomy with Neural Environment Representations

, Associate Professor of Aeronautics and Astronautics, Stanford University
New developments in computer vision and deep learning powered by next-generation GPUs have led to the rise of neural environment representations: 3D maps that are stored as deep networks that spatially register occupancy, color, texture, and other physical properties. These environment models can generate photorealistic synthetic images from unseen viewpoints, and can store 3D information in exquisite detail. I'll investigate the question: How can robots use neural environment representations for perception, motion planning, manipulation, and simulation? I'll show recent work from my lab in which we use neural radiance field (NeRF) representations for robot navigation and manipulation, and incorporate the NeRF model into a differentiable robot dynamics simulator. I'll conclude with future opportunities and challenges in integrating neural environment representations into the robot autonomy stack.
活动: GTC Digital September
日期: September 2022
行业: 所有行业
话题: Autonomous Machines - Robotics
级别: 中级技术
语言: 英语
话题: Autonomous Machines - Robotics Research
所在地: