The SAIL-Toyota Center for AI research hosted the 5th annual Toyota Research Institute Joint University Workshop on Jan 13 and 14. Bringing together faculty, students, and staff from Stanford, The University of Michigan, and MIT as well as TRI leadership and research staff the two-day conference featured talks and posters on the work being done with TRI's partnership with the three universities.

The talks featured during the workshop included:

• Safe Feedback Interactions in Human-Autonomous Vehicle Systems, Chris Gerdes (Stanford University)
• Program Synthesis for the Physical World, Armando Solar-Lezama (MIT)
• Manipulation in (Partially) Occluded Environments, Dmitry Berenson (University of Michigan)
• Development of a "Primary Other Vehicle" for the Testing and Evaluation of Highly Automated Vehicles, Huie Peng (University of Michigan)
• Highly Dynamic Bipedal Locomotion in Unknown, Loosely Structured Environments, Maani Ghaffari (University of Michigan)
• Deep Orientation Uncertainty Learning, Igor Gilitschenski (MIT)
• Risk-Aware and Socially Aware Autonomy, Alyssa Pierson (MIT)
• Joint Contextual Forecasting of Vehicle Behaviors, Actions, and Trajectories, Juan Carlos Niebles (Stanford University)
• Validation of Human-Autonomous Vehicle Systems, Mykel Kochenderfer (Stanford University)
• Formally Verified Guardians to Avoid Moving Obstacles, Jean Baptiste Jeannin (University of Michigan)
• Safety Interlock for Self-driving Cars, Valerie Richmond and Jeff Chow (MIT)
• Collision Avoidance Guardian at the Dynamic Limits of Handling, Tulga Ersal (University of Michigan)
• Soft Robotics and Haptics for Shared Autonomy and Teleoperation, Allison Okamura (Stanford University)
• Acquiring, Representing, and Transporting Manipulation Knowledge, Jeannette Bohg (Stanford University)
• Interactive Learning for Manipulation Piles of Stuff, Brent Griffin (University of Michigan)
• Robotic Manipulation Data Engine, Siyuan Dong (MIT)
• Dense, Freeform Tactile Feedback for Manipulation and Control, Wojciech Matusik (MIT)
• VehiQl - Query Processing for Visual Data Streams, Tom Wenisch (University of Michigan)Sensible Deep Learning for 3D Data, Justin Solomon (MIT)
• Supersizing Robot Learning Through Fast Imitation, Roberto Martin-Martin (Stanford University)
• Building Fully 3D Representations, David Fouhey (University of Michigan)
• Camera-based Proprioceptions and Tactile Sensing, Edward Adelson (MIT)
• It's the Transitions: Supporting Shared Control in Vehicle Steering Across Routine and Off-Nominal Conditions, Brent Gillespie (University of Michigan)
• Driver Perception and the Car-to-Driver Handoff, Benjamin Wolfe (MIT)
• Human Behaviors and interaction for In-Car Experiences, Michael Bernstein (Stanford University)
• Individual Driving Styles: Learning, Modeling, and Improving Driver Safety, David Sirkin (Stanford University)
• Machines That Can Introspect, Andrei Barbu (MIT)
• Analysis by Synthesis Revisited: Visual Scene Understanding by Integrating Probabilistic Programs and Deep Learning, Jiajun Wu and Josh Tenenbaum (MIT)
• Using Deep Learning to Speed Up Deep Learning, Fredo Durand (MIT)
• Multimodal Sensing of Human Behavior, Emily Provost (University of Michigan)
• Data-Efficient Algorithms for Machine Learning with Provable Guarantee, Tengyu Ma (Stanford University)
• Developing A Personalized Guardian System to Assist Aging Drivers Through Machine Learning, Sensor Fusion, and Data Mining, Yi Lu Murphey (University of Michigan)

For more photos from the TRI Joint University Workshop visit our event gallery