Announcing an AI Meetup with Coline Devin, from UC Berkeley. She will be talking about her recent work on robot perception.
Note: Meetup has been moved to Monday 10/9
Bio: Coline is a PhD student in BAIR at UC Berkeley, advised by Professors Pieter Abbeel and Trevor Darrell. She graduated in Computer Science from Harvey Mudd College in May 2015, and has previously interned at Google Deepmind and MILA.
Talk Title: Deep Object-Centric Representations for Generalizable Robot Learning
The talk will be primarily centered around this paper: https://arxiv.org/pdf/1708.04225.pdf
Robotic manipulation in complex open-world scenarios requires both reliable physical manipulation skills and effective and generalizable perception. In this paper, we propose a method where general purpose pretrained visual models serve as an object-centric prior for the perception system of a learned policy. We devise an object-level attentional mechanism that can be used to determine relevant objects from a few trajectories or demonstrations, and then immediately incorporate those objects into a learned policy. A task-independent meta-attention locates possible objects in the scene, and a task-specific attention identifies which objects are predictive of the trajectories. The scope of the task-specific attention is easily adjusted by showing demonstrations with distractor objects or with diverse relevant objects. Our results indicate that this approach exhibits good generalization across object instances using very few samples, and can be used to learn a variety of manipulation tasks using reinforcement learning.
Claim the event and start manage its content.I am the organizer