Mariah Schrum
InterACT Lab
Berkeley Artificial Intelligence Research (BAIR)
UC Berkeley
Berkeley Way West, Berkeley, CA 94704
I am currently a postdoc at UC Berkeley working with Anca Dragan in the InterACT Lab. In 2023, I completed my PhD in Robotics at the Georgia Institute of Technology, working with Matthew Gombolay in the CORE Robotics Lab. In 2020, I received my Master's in Computer Science from Georgia Tech and my bachelor's in Biomedical Engineering in 2018 from Johns Hopkins. I am a recepient of the Accessibility, Rehabilitation, and Movement Science Fellowship.
I currently work on Reinforcement Learning applications in the real world, with a focus on RL for Deep Brain Stimulation. During my PhD, I worked on deep learning algorithms for personalizing autonomous systems to account for heterogeneity in human-machine interaction. My belief is that robots and AI systems must not only be designed to effectively perform the tasks for which they were intended, but they must also be optimized to work alongside humans. In 2022, I interned at Toyota Research Institute, investigating a data-driven approach for optimizing autonomous vehicle driving style. In 2021, I interned at Inuitive Surgical and worked on their new Ion surgical robot.
news
Our paper, Coprocessor Actor Critic: A Model-Based Reinforcement Learning Approach For Adaptive Brain Stimulation, has been accepted to International Conference on Machine Learning in Vienna!
Our paper, MAVERIC: A Data-Driven Approach to Personalized Autonomous Driving, has been accepted to IEEE Transactions on Robotics and will be presented at IROS in Abu Dhabi!
Our paper, Mixed-Initiative Multiagent Apprenticeship Learning for Human Training of Robot Teams, has been accepted at Neurips '23!
I started my postdoc at UC Berkeley with Anca Dragan in the InterACT Lab!
I am honored to have been selected to participated in RSS Pioneers '23!
Our paper, Investigating the Impact of Experience on a User's Ability to Perform Hierarchical Abstraction, is nominated for Best Student Paper at RSS '23!
I gave an invited talk on my research for the Semiautonomous seminar series at UC Berkeley.
I was invited to present on my work at Georgia Tech's CS 7633: Human-Robot Interaction graduate course.
My poster on our work MAVERIC: A Data-Driven Approach to Personalized Autonomous Driving won best poster at the CRIDC poster competition.
I gave an invited talk on my research for the Robotics Seminar series at the Colorado School of Mines.
I presented on our work, Reciprocal MIND MELD: Improving Learning From Demonstration via Personalized, Reciprocal Teaching , at CoRL in New Zealand!.
I was invited to present on my work at University of Utah's CS 6960: Human-AI Alignment graduate course.
Our review paper on best practices in HRI, Concerning Trends in Likert Scale Usage in Human-Robot Interaction: Towards Improving Best Practices , has been accepted to Transactions on Human-Robot Interaction (tHRI).
I presented on our work, Meta-Active Learning in Probabilistically Safe Optimization , at IROS in Japan!
Our paper Reciprocal MIND MELD: Improving Learning From Demonstration via Personalized, Reciprocal Teaching has been accepted to the Conference on Robot Learning (CoRL).
I gave an invited talk on my work at the Mines Interactive Robotics Research Lab Summer Speaker Series .
I’ll be joining Toyota Research Institute as a summer research intern!
I’ve officially passed my PhD thesis proposal titled “Data-Driven Personalization Techniques to Account for Heterogeneity in Human-Robot Interaction” and am now a Ph.D. candidate!
Our poster on our work MIND MELD: Personalized meta-learning for robot-centric imitation learning won best poster at the IRIM Research Showcase.
I gave an invited talk on my work at Affective Intelligence and Robotics Laboratory in Cambridge, UK!.
Our paper MIND MELD: Personalized meta-learning for robot-centric imitation learning won best Technical Paper at the Conference on Human-Robot Interaction!.
I gave an invited talk on our work , MIND MELD: Personalized meta-learning for robot-centric imitation learning , at the Workshop on Machine Learning in Human-Robot Collaboration: Bridging the Gap .
I gave an invited talk on our work, MIND MELD: Personalized meta-learning for robot-centric imitation learning , at the Workshop on Human-Robot Interactive Learning .
I am grateful to be part of the HRI Pioneers 2022 cohort and to participate in this great workshop!.
Our poster on our work MIND MELD: Personalized meta-learning for robot-centric imitation learning won best poster at the CRIDC poster competition.