Z Advanced Computing (ZAC) has received funding from the US Air Force (USAF) to develop 3D image recognition technology for unmanned aerial vehicles (UAV).
The company’s drone image recognition technology is based on explainable artificial intelligence (AI). ZAC will work on developing image recognition technology for drones to meet the USAF’s aerial image / object recognition requirements.
The start-up demonstrated its explainable AI (XAI) capability, which allows recognition of details of 3D objects from any direction.
ZAC CTO Dr Saied Tadayon said: “With our superior approach, complex 3D objects can be recognised from any direction, using only a small number of training samples.
“You cannot do this with the other techniques such as deep convolutional neural networks (CNNs), even with an extremely large number of training samples. That’s basically hitting the limits of the CNNs.”
According to the Defense Advanced Research Projects Agency (DARPA), XAI will help future troops to understand an ’emerging generation of artificially intelligent machine partners’.
Stressing on the need for developing XAI, DARPA stated that the autonomous systems developed using machine learning models pose challenges for humans as they are difficult to understand.
Dr Saied Tadayon added: “For complex tasks such as drone vision you need ZAC’s superior technology to handle detailed 3D image recognition.”
ZAC’s technology has the ability to allow image search to be performed in several applications, including e-commerce, ad network, autonomous vehicles, medical imaging, satellite imaging, face recognition, video, and security.
In May, the USAF signed an agreement with Massachusetts Institute of Technology (MIT) to start an AI accelerator programme to improve airforce operations and deliver innovations in disaster response and medical readiness.