News/Research

UC Berkeley at Conference on Robotic Learning

14 Sep, 2017

UC Berkeley at Conference on Robotic Learning

An amazing showing for UC Berkeley at the first annual Conference of Robotic Learning this November, with 14 papers published by UC Berkeley authors, 3 co-authored by BCNM's own Ken Goldberg! The conference aims to gather approximately 250 of the best researchers currently studying robotics and machine learning and discuss advances in the field.

Goldberg's research papers — "Dart: Optimizing Noise Injection in Imitation Learning," "DDCO: Discovery of Deep Continuous Options for Robot Learning from Demonstrations," and "Learning Deep Policies for Robot Bin Picking using Discrete-Event Simulation of Robust Grasping Sequences" — will be published through the JMLR Workshop & Conference Proceedings series and presented at the conference.