Stanford Seminar - Representation Learning for Autonomous Robots, Anima Anandkumar

Autonomous robots need to be efficient and agile, and be able to handle a wide range of tasks and environmental conditions. This requires the ability to learn good representations of domains and tasks using a variety of sources such as demonstrations and simulations. Representation learning for robotic tasks needs to be generalizable and robust.

I will describe some key ingredients to enable this: (1) robust self-supervised learning (2) uncertainty awareness (3) compositionality. We utilize NVIDIA Isaac for GPU-accelerated robot learning at scale on a variety of tasks and domains.

Generlizable Learning for Robotics.
Trinity of Generalizable AI.
Physical World if Continuous.
Motivating Problem in Robotics.
State estimation through PDE observer.
Grid-free learning for continuous phenomena.
Neural Operator.
Fourier Transform for global convolution.
FNO: Fourier Neural Operator.
First ML method to solve fluid flow.
Nvidia Modulus.
Operational Space Control (OSC).
Reducing Supervision and enhancing robustness.
Conclusion.