Multimodal Human Identification for Computer Security
Abstract
(A) A cooperative coevolutionary approach for object detection is developed. It fuses the scene contextual information with the available statistical and prediction information available from color and infrared sensors. The sensor fusion system maintains high detection rates under a variety of environmental conditions. The results are shown for a full 24 hour diurnal cycle. (b) An agent-based intrusion detection system, where evolutionary computational techniques, similar to those discussed in (a) are explored. A detailed architecture for a coevolutionary agent based system is given and the concept of super agent is described. (c) A performance modeling approach for object recognition is developed and the results are shown on synthetic aperture radar images.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 15, 2005
- Accession Number
- ADA430881
Entities
People
- Bir Bhanu
- Edward Hong
- Sohail Nadimi
Organizations
- University of California, Riverside