The U.S. Air Force (USAF) was so impressed by the brief time and rapid learning process shown during a demonstration of three-dimensional (3D) image-recognition software, it awarded a second contract to startup software company Z Advanced Computing, Inc. The company is a pioneer in the development of cognitive explainable artificial intelligence (Cognitive XAI) software, which combines artificial intelligence (AI) and machine learning (ML) techniques for 3D object/image recognition from any angle. The software is effective using only a few training samples and without need of an overly powerful laptop computer for data processing.
The short training cycle for the software is in contrast to many object/image recognition programs that run on large servers and require thousands or even millions to billions of training samples for effective 3D object recognition, especially at different angles. Dr. Bijan Tadayon, the chief executive officer (CEO) of ZAC, said: “This cannot be done with the other algorithms, such as Deep Convolutional Neural Networks (CNN) or ResNets, even with an extremely large number of training samples or a large number of GPU servers.” Dr. Saied Tadayon, Chief Technology Officer (CTO) of ZAC, added: “Deep CNN is very fragile and is easily fooled, and its output is not backed up by explanation.” The Cognitive XAI software has a wide range of expected applications, including for autonomous vehicles, medical systems, satellite/aerial imaging, and security systems.