baiosphere’s Post

View organization page for baiosphere, graphic

2,496 followers

Two down, three more to go in our Munich AI Lectures summer sprint! Last week, Ludovic Righetti from New York University highlighted the innovative integration of Model-Predictive Control (MPC) and Reinforcement Learning (RL) in robotics. He explained how MPC offers precise control while RL provides adaptability, creating a robust framework for complex robotic tasks. Combining these approaches can overcome individual limitations and enhance robotic motion and manipulation. Prof. Righetti also emphasized interdisciplinary openness and ethical considerations for AI and robotics researchers.   Yesterday, Ivan Laptev from MBZUAI (Mohamed bin Zayed University of Artificial Intelligence) discussed the evolution and current challenges in image and video recognition. He highlighted how human pose estimation and object recognition have become standard, yet image recognition software can miss context if queried incorrectly. Prof. Laptev introduced ViViDex, a framework for teaching robots precise hand movements using human videos. ViViDex involves three steps: extracting human movements, teaching robots through these movements, and refining their skills using video data. This research significantly advances robotic capabilities for complex tasks.   These lectures underscore the need to view robotics and AI from diverse perspectives, combining insights from various researchers and disciplines to make meaningful progress. We are excited to see what Alexei (A.) Efros will add to the conversation in his highlight lecture ‘We are (still?) not giving data enough credit’ on July 17th - Sign up now! https://lnkd.in/eFH5iNJU #AI #computerVision #robotics #munichAILectures #baiosphere

  • No alternative text description for this image
  • No alternative text description for this image
  • No alternative text description for this image

To view or add a comment, sign in

Explore topics