
(4) Deep Defense of Distributed Deep Learning (D4L)Award last edited on: 5/5/2023
Sponsored Program
SBIRAwarding Agency
DOD : NavyTotal Award Amount
$139,974Award Phase
1Solicitation Topic Code
N193-A01Principal Investigator
Charles O'NeilCompany Information
Phase I
Contract Number: N68335-20-F-0169Start Date: 11/21/2019 Completed: 4/20/2020
Phase I year
2020Phase I Amount
$139,974Benefit:
All 9 topics under the Navy Technology Acceleration objectives for AI/ML would benefit from distributed deep learning (DDL) defense schemes: (1) Manned and unmanned aircraft could employ a concurrent, divide-and-combine DDL computation model to enable real-time learning of big data using a deep defense strategy. (2) A ship fleet could collaborate using secure pattern learning of surrounding objects, weather, and threats. (3) Naval Special Warfare operators could employ secure wireless communications under enemy jamming attack. (4) Naval satellite sensors could employ localized processing and global fusion to detect intrusion.
Keywords:
Ensemble Bayesian Learning, Ensemble Bayesian Learning, algorithm, Blockchain, ensemble, Distributed Deep Learning, AGBS, spiral attack
Phase II
Contract Number: ----------Start Date: 00/00/00 Completed: 00/00/00