Real-Time Image Detection w/ ARKit

About us: We're a venture-backed company in North America with an integrated team of designers and developers working at the intersection of machine learning and computer vision to fundamentally revolutionize how people interact with one another. Objective: We are looking for an AR Kit and graphics expert who might be able to help on a contract basis for doing live-image registration of an object from different angles on the iPhone. Details: This person needs to be able to answer questions such as... - How close does the live image you're looking at with the device have to match the reference image? (Things like, if I can see 80% of the reference image does it fire? Or does it require all 100% of the image in order to match? As well as things like how much of an angle can I be looking at a matching image, does it have to be directly square on to match or can I be to the side by a certain angle?) - How well the recognized image position stays in place, and how well it positions the recognized image plane? - Can you explain how the physical size property of the reference image is used to assist with image recognition? - Can you explain how lighting and/or device tracking affects the image recognition?