Let's reverse engineer the phenomenal Tesla Optimus. No insider info, just my own analysis. Long read: 1. The smooth hand movements are almost certainly trained by imitation learning ("behavior cloning") from human operators. The alternative is reinforcement learning in

Solid work!

Heng Wang and Cordelia Schmid win the test of time award for 'Action recognition with improved trajectories'!

🚀Presenting FeatureNeRF at #ICCV2023! We propose a unified framework for 3D representation learning, distilling 2D foundation models into generalizable 3D semantic representations via neural rendering. 🗓️Join our Thursday Morning poster session at Room Nord 074. 🧵(1/n)

🚨 Announcing TD-MPC2: Scalable, Robust World Models for Continuous Control TD-MPC2 performs 100+ tasks without tuning, and allows us to train a single 317M parameter model to perform 80 tasks across multiple domains, embodiments, and action spaces! nicklashansen.github.io/td-mpc2 🧵1/

Let’s think about humanoid robots outside carrying the box. How about having the humanoid come out the door, interact with humans, and even dance? Introducing Expressive Whole-Body Control for Humanoid Robots: expressive-humanoid.github.io See how our robot performs rich, diverse,

Want to obtain time-consistent dynamic mesh from monocular videos? Introducing: Dynamic Gaussians Mesh: Consistent Mesh Reconstruction from Monocular Videos liuisabella.com/DG-Mesh/ We reconstruct meshes with flexible topology change and build the corresp. across meshes. 🧵(1/n)

Congratulations Dr. Yuzhe Qin Yuzhe Qin and Dr. Yueh-Hua Wu Kris Wu!

Can RL take the F1 championship from Max one day? Cool work by Adrian Remonda, Nicklas Hansen and team.

Meet our AI-powered robot that’s ready to play table tennis. 🤖🏓 It’s the first agent to achieve amateur human level performance in this sport. Here’s how it works. 🧵

Super impressive work on fine-tuning IL policies with RL using sparse rewards!

🤖 Want your robot to grab you a drink from the kitchen downstairs? 🚀 Introducing BUMBLE: a framework to solve building-wide mobile manipulation tasks by harnessing the power of Vision-Language Models (VLMs). 👇 (1/5) 🌐 robin-lab.cs.utexas.edu/BUMBLE

Shivin and team show how to turn data curation for behavior cloning into a supervised learning problem using DataModels. Clean formulation and strong results!