MSR Thesis Presentation
Learning Dexterous Manipulation from Grasp Priors and Force-Aware Demonstrations
Abstract: Dexterous manipulation requires coordinated finger motion, stable hand–object contact, and appropriate force application. Learning these behaviors from demonstrations is challenging because task-specific robot data is costly to collect and motion commands alone do not always produce reliable contact. This thesis studies two complementary approaches to improving the data used for dexterous policy learning: large-scale [...]
Toward Fully Autonomous Agile Flight under Partial Observability
Abstract: This thesis explores sensing and acting as two complementary components for autonomous agile flight under partial observability. For sensing, monocular depth predictions are rescaled using sparse metric features from visual-inertial navigation, enabling collision avoidance without a prebuilt map or dedicated depth sensor. For acting, a recurrent policy is trained through a differentiable nonlinear quadrotor [...]