Dynamic data driven applications systems with multi-modal sensing, collaborative perception and deep computing
Abstract
The research and development conducted by both PIs, Profs. Wei and Zhu, have been consistently supported by ARO, AFOSR and AFRL. One fundamental question that comes up from our productive collaborations is: what is a principled way to integrate multimodal sensing and deep computing in a collaborative manner, from sensing to classification and detection? This will be the focus of this proposal with three important thrusts: 1) Sensor Phenomenology: Design experiments emulating ISR missions and collect big volume of data from the multimodal sensors and conduct rigorous exploratory analytics to extract their inherent signatures for improving detection/identification accuracy. 2) Collaborative Perception: To systematically study the collaborations among multi-modal sensory data from different spatial distance/resolution and frequency spectrum so they can be integrated to provide complementary information and thus forming a more complete picture of the targets or regions of interest. 3) Deep Computing: To explore and examine effective data representations and algorithms by using cutting-edge computing methods from statistical signal processing, high-performance computing, machine learning (esp. deep reinforcement learning and transfer learning).
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 21, 2022
- Source ID
- FA95502110082XX0
Entities
People
- Jie Wei
Organizations
- Air Force Office of Scientific Research
- Research Foundation of The City University of New York
- United States Air Force