Detecting Anomalies in Object Appearance and Motion Dynamics

Graduate Project

Pubblico Deposited

Citeable URL: https://ir.library.oregonstate.edu/concern/graduate_projects/6d5704477

Descriptions

Attribute Name	Values
Creator	Alotaibi, Mazen
Abstract	Machine common sense remains a broad, potentially unbounded problem in AI. Our focus is to move toward AI systems that can develop common-sense reasoning similar to humans to detect anomalies. In particular, we study the problem of detecting the violation of expectations when object appearance or motion dynamics change from simulated experiments. We have developed a system of multiple components to solve the problem of machine common sense. The system contains three main components: a two-stage tracker, a role assigner, and a rule-based reasoning agent. We evaluate our system on scenes from the DARPA Machine Common Sense, Passive Violation of Expectation task.
Contributor	Kim, Chanho Shureih, Zeyad Malik, Ashish Irvine, Jed Lee, Stefan Li, Fuxin Fern, Alan
Resource Type	Capstone Project
Date Issued	2022-08-29
Degree Level	Master's
Degree Name	Master of Science (M.S.)
Degree Field	Artificial Intelligence
Degree Grantor	Oregon State University
Commencement Year	2023
Advisor	Li, Fuxin
Committee Member	Fern, Alan Lee, Stefan
Academic Affiliation	Electrical Engineering and Computer Science
Dichiarazione dei diritti	In Copyright
Publisher	Oregon State University
Peer Reviewed	No
Language	English [eng]

Thumbnail	Titolo	Data caricata	Visibilità	Azioni
	mazen_alotaibi_2022.pdf	2022-09-07	Pubblico	Scaricare