- Patent Title: Systems and methods enabling online one-shot learning and generalization by intelligent systems of task-relevant features and transfer to a cohort of intelligent systems
-
Application No.: US16098785Application Date: 2017-05-10
-
Publication No.: US11645444B2Publication Date: 2023-05-09
- Inventor: Matthias J. Scheutz , Evan A. Krause
- Applicant: Trustees of Tufts College
- Applicant Address: US MA Medford
- Assignee: Trustees of Tufts College
- Current Assignee: Trustees of Tufts College
- Current Assignee Address: US MA Medford
- Agency: Occhiuti & Rohlicek LLP
- International Application: PCT/US2017/032010 2017.05.10
- International Announcement: WO2017/197018A 2017.11.16
- Date entered country: 2018-11-02
- Main IPC: G06F40/00
- IPC: G06F40/00 ; G06N5/02 ; G10L15/18 ; G06F16/33 ; G06F16/9032 ; G06F16/903

Abstract:
An intelligent system, such as an autonomous robot agent, includes systems and methods to learn various aspects about a task in response to instructions received from a human instructor, to apply the instructed knowledge immediately during task performance following the instruction, and to instruct other intelligent systems about the knowledge for performing the task. The learning is accomplished free of training the intelligent system. The instructions from the human instructor may be provided in a natural language format and may include deictic references. The instructions may be received while the intelligent system is online, and may be provided to the intelligent system in one shot, e.g., in a single encounter or transaction with the human instructor.
Information query