IEI holds the basic patents for allowing large assemblies of artificial neural networks to connect themselves into complex neural architectures that emulate the visual pathways of the brain. Using  this extremely valuable form of autonomous brain building, our company excels beyond all others in building systems that (1)  can detect anomalies, even within dynamic scenes, and (2) readily classify objects and scenarios independently of scale and viewing angle. In both of these cases, the need for rule-based programming is greatly minimized and finished products may be realized in a small fraction of the time and expense required of conventional machine vision applications, neural network-based or otherwise.

In the figure to right, for instance, our SuperNet-based application has shown its worth at autonomous identification and targeting of various military aircraft. Rather than expend hundreds of thousands of dollars for human programmers to develop template-based computer programs that recognize distinguishing features of various enemy aircraft, our systems simply require inexpensive cameras sampling the target from various aspect angles. Rather than require months or years of software development, our systems self-organize themselves over just seconds.

Currently IEI machine vision systems are being utilized by a major automotive parts manufacturer for intelligent headlight control and occupant identification. The Transportation Safety Administration (TSA) is currently contracting with our company to build smart sensor networks involving our machine vision applications to detect various kinds of vehicular traffic anomalies, ramp and runway threats, and to limit vehicular access within secured areas of the airport. Complimenting the TSA effort, the Air Force's SoftNet program is now using IEI's image anomaly detection system to detect various kinds of anomalies taking place on aircraft runways. Recently, IEI has licensed many features of its machine vision capabilities to the automotive industry to address a variety of vehicular safety issues.

Our company has also planted the flag in the area by patenting the key neural network methodologies for developing plans of action (i.e., best way to fire a missile or acquire an object in a factory setting) based upon live video streams or any other kind of sensor input. Such context-dependent Creativity Machines have excelled where contemplation of the environment is necessary to interpret things and activities therein and to deliberate upon an optimal course of action.

IEI builds tailored machine vision applications for government and industry. The company likewise engages in various joint ventures in advanced machine vision.