Kick off of 2023-2024 Seminar Series

We're excited to kick off another season of our virtual seminar series for the TC on Model-Based Optimization for Robotics.

We'll get rolling this week with a talk from Yuval Tassa of Google DeepMind at 10AM EST on Friday (1st December 2023). Please find the flyer with all the details at this link.

Speaker: Yuval Tassa (Google DeepMind)

Title: MJPC: Asynchronous UI for real-time behavior synthesis

Abstract:

Model-free, learning-based methods are simple and powerful but inherently slow. Model-based predictive control methods can synthesize behavior in real time, but are considered difficult to understand, implement and use.

In this talk I will present MuJoCo MPC (MJPC), a fully open-source interactive application and software framework for predictive control that lets the user easily author and solve complex robotics tasks. MJPC is a rich GUI application, giving the user real-time visual feedback and allowing them to develop intuitions about the underlying algorithm. Importantly, the interactive simulation can be slowed-down asynchronously—effectively speeding up the controller—leading to a democratization of predictive control tooling.

In closing, I will discuss various ways in which model-based and model-free methods can be combined, exploiting the strengths of both approaches.

The talk will include a live demo of multiple locomotion and manipulation tasks solved from scratch, in real-time, on a laptop.

Date: Friday, December 1st 2023

Time: 10:00-11:00 AM EST (GMT -05:00)

Link: https://columbiauniversity.zoom.us/j/91247893326?pwd=L2JWU21aQzc4cU1ZQklEb0QrWGQvdz09

More details on upcoming seminars (and video links for past ones) can be found here.