Proceedings

Panel

Panel Questions

Please post your questions and any discussion in this slack channel by June 10 for inclusion in our panel discussion with invited speakers (joining instructions are available after registering here or via the ICRA slack). Our panel will consist of:

• Nathan Ratliff (NVIDIA)
• Aude Billard (EPFL)
• Sergey Levine (UC Berkeley)
• Michael Beetz (University Bremen)
• Marc Toussaint (TU Berline)
• Emre Ugur (Bogazici University)

Invited Talks

Sergey Levine (UC Berkeley), Model-based RL

Slack Q&A

Emre Ugur (Bogazici University), Predictive Robotic Manipulation

Slack Q&A

Nathan Ratliff (NVIDIA) , Optimization Over a Geometric Fabric

Slack Q&A

Michael Beetz (University Bremen), Learning human scale manipulation tasks

Slack Q&A

Accepted Papers and Talks

All the papers are available in this Google Drive, the talks are on this Youtube Playlist, and the discussion can be found at this Slack Workspace (joining instructions are available after registering here or via the ICRA slack). Paper Q&As are available in channels ws20_01, ws20_02, ..., ws20_17.

Closed-loop behaviour of learned models in robotic food-cutting

by Ioanna Mitsioni (KTH); Yiannis Karayiannidis (Chalmers University of Technology); Danica Kragic (KTH)

Presentation, Paper, Q&A

Learning Visuomotor Policies for Aerial Navigation Using Cross-Modal Representations

by Rogerio Bonatti (CMU); Ratnesh Madaan (Microsoft); Vibhav Vineet (Microsoft Research); Sebastian Scherer (CMU); Ashish Kapoor (Microsoft)

Presentation, Paper, Q&A

Hierarchical Foresight: Self-Supervised Learning of Long-Horizon Tasks via Visual Subgoal Generation

by Suraj Nair (Stanford University); Chelsea Finn (Google Brain)

Poster. Paper, Q&A

Time-Informed Exploration For Robot Motion Planning

by Sagar S Joshi (Georgia Institute of Technology); Seth Hutchinson (Georgia Tech); Panagiotis Tsiotras (Georgia Institute of Technology)

Presentation, Paper, Q&A

Hindsight for Foresight: Unsupervised Structured Dynamics Models from Physical Interaction

by Iman Nematollahi (University of Freiburg); Oier Mees (University of Freiburg); Lukas Hermann (University of Freiburg); Wolfram Burgard (University of Freiburg)

Presentation, Paper, Q&A

Learning to Fly via Deep Model-Based Reinforcement Learning

by Philip Becker-Ehmck (Volkswagen Group); Maximilian Karl (Machine Learning Research Lab, Volkswagen Group); Jan Peters (TU Darmstadt); Patrick van der Smagt Volkswagen Group)

Presentation, Paper, Q&A

DISCO: Double Likelihood-free Inference Stochastic Control

by Lucas Barcelos (The University of Sydney); Rafael Oliveira (The University of Sydney); Rafael Possas (University of Sydney); Lionel Ott (The University of Sydney); Fabio Ramos (NVIDIA, The University of Sydney)

Presentation, Paper, Q&A

APPLD: Adaptive Planner Parameter Learning from Demonstration

by Xuesu Xiao (UT Austin); Bo Liu (UT Austin); Garrett Warnell (US Army Research Laboratory); Jonathan Fink (US Army Research Lab); Peter Stone (UT Austin)

Presentation, Paper, Q&A

Lazy Action Value Models for Q-Learning in Robotics

by Ingmar F Schubert (TU Berlin); Marc Toussaint (TU Berlin and Max Planck)

Presentation, Paper, Q&A

Visual Navigation Among Humans With Optimal Control as A Supervisor

by Varun Tolani (UC Berkeley); Somil Bansal (UC Berkeley); Aleksandra Faust (Google Brain); Claire Tomlin (UC Berkeley)

Presentation, Paper, Q&A

Stein Variational Model Predictive Control

by Alexander Lambert (Georgia Institute of Technology); Adam Fishman (U Washington); Dieter Fox (NVIDIA Research / U Washington); Byron Boots (U Washington); Fabio Ramos (U Sydney / NVIDIA)

Presentation, Paper, Q&A

Data-efficient Control from Images by Learning How to Use a Simple Model

by Thomas J Power (U Michigan); Dmitry Berenson (U Michigan)

Presentation, Paper, Q&A

Deep Visual Reasoning: Learning to Predict Action Sequences for Task and Motion Planning from an Initial Scene Image

by Danny Driess (University of Stuttgart); Jung-Su Ha (Jung-Su Ha); Marc Toussaint (TU Berlin)

Presentation, Paper, Q&A

Robotic Motion Planning using Learned Critical Sources and Local Sampling

by Rajat K Jenamani (IIT Kharagpur); Rahul Kumar (IIT Kharagpur); Parth Mall (IIT Kharagpur); Kushal Kedia (IIT Kharagpur)

Presentation, Paper, Q&A

Learning When to Trust a Dynamics Model When Planning With Physical Constraints

by Peter Mitrano (U Michigan); Dale McConachie (U Michigan); Dmitry Berenson (U Michigan)

Presentation, Paper, Q&A

Information Theoretic Model Predictive Q-Learning

by Mohak Bhardwaj (U Washington); Ankur Handa (NVIDIA); Dieter Fox (NVIDIA Research / U Washington); Byron Boots (U Washington)

Paper. Q&A

Learning Contextual Actions for Heuristic Search-Based Motion Planning

by Dhruv Mauria Saxena (CMU); and Maxim Likhachev (CMU)

Paper, Q&A