PERISCOPE: Perceptual Representations for Actions, Composition, and Verification
Abstract
In this project we will develop perception-in-the-loop abstraction, speci?cations, ver-i?cation, composition, and repair techniques" that integrate perception and action ofautonomous systems operating in unstructured and changing environments. Current autonomous"systems can perform impressive tasks, from retrieving target objects in clutter toying at high speed through a forest; however, no e""xisting autonomous system can perform ahigh-level task over an extended period of time in an unstructured and changing environment,""let alone provide guarantees for such behavior. Over time, prior assumptions encoded intraining data and model structure are viola""ted, inevitably resulting in system failure.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 07, 2017
- Source ID
- N000141712699
Entities
People
- Hadas Kress-Gazit
Organizations
- Cornell University
- Office of Naval Research
- United States Navy