Target cuing: a heterogeneous neural network approach

摘要

Abstract: Autonomous analysis of complex image data is a criticaltechnology in today's world of expanding automation.The growth of this critical field is slowed by problemsin traditional image analysis methods. Traditionalmethods lack the speed, generality, and robustness thatmany modern image analysis problems require. Whileneural networks promise to improve traditionaltechniques, homogeneous neural network systems havedifficulty performing all the diverse analysis requiredof an autonomous system. This paper proposes adual-staged, heterogeneous neural network approach toimage analysis; specifically, a way to solve the targetcuing problem.!