Affordance Learning Learning Affordances from Human Videos for Autonomous Robotic Exploration
Olivia Y. Lee, Annie Xie, Karl Pertsch, Suraj Nair, and Chelsea Finn
Ongoing Project, presented at the Stanford Symbolic Systems Annual Research Showcase
We are developing a real robot system that autonomously explores and gathers in-domain experience by leveraging priors from human data. To do so, we address the lack of well-shaped rewards, a key challenge for robotic exploration, by leveraging structural assumptions from human interaction data. I have conducted preliminary experiments in simulation and am currently testing the approach on a real robot system.

Play it by Ear Play it by Ear: Learning Skills amidst Occlusion through Audio-Visual Imitation Learning [paper] [site] [publication]
Maximilian Du*, Olivia Y. Lee*, Suraj Nair, and Chelsea Finn
Robotics Science and Systems (RSS), 2022 [recording]
We propose a system that learns to complete challenging, partially-observed manipulation tasks by reasoning over visual and audio inputs. Our system combines offline imitation learning from a modest number of tele-operated demonstrations and online finetuning using human provided interventions. In simulation, our system benefits from using audio and online interventions improve the success rate of offline imitation learning by ~20%. On a Franka Emika Panda robot, our system completes manipulation tasks (e.g. extracting keys from a bag) with a 70% success rate, 50% higher than a policy that does not use audio.

An updated analysis of satellite quantum-key distribution missions [paper]
Olivia Y. Lee, Tom Vergoossen
We provide a meta-analysis of research and development efforts that have contributed to the realization of satellite quantum-key distribution (satellite-QKD). We present an overview of the various technical parameters of performing satellite-based QKD regarding protocols and infrastructure. Following a timeline of key milestones in the development of satellite-QKD, we present a high-level summary of globally completed and proposed quantum satellite missions, categorized by specific advancements thus far in satellite-QKD. We conclude with a discussion on the technical challenges currently faced in satellite-QKD and future directions to address these challenges.