Apple has been granted a patent (number 11468275 B1) for “computer vision using a prior probability distribution selected based on an image capture condition.” It’s for improving the iPad’s ability to scan and capture images.
About the patent
In the patent Apple says that, like human vision, computer vision tasks generally attempt to reason from one or more images. Humans, however, frequently have the benefit of at least some background knowledge or expectations that are useful to such reasoning.
For example, humans often have implicit notions of where certain types of things might be in the areas around them. As a specific example, a human, even with eyes closed, may have a notion that a table is more likely to be ahead of him than above him on the ceiling.
Computer vision tasks, in contrast, generally attempt to reason from images without the benefit of such background knowledge and expectations and thus can be less accurate, efficient, or effective than desired.
Summary of the patent
Here’s Apple’s abstract of the patent: “A machine learning (ML) model is trained and used to produce a probability distribution associated with a computer vision task. The ML model uses a prior probability distribution associated with a particular image capture condition determined based on sensor data. For example, given that an image was captured by an image capture device at a particular height above the floor and angle relative to the vertical world axis, a prior probability distribution for that particular image capture device condition can be used in performing a computer vision task on the image.
“Accordingly, the machine learning model is given the image as input as well as the prior probability distribution for the particular image capture device condition. The use of the prior probability distribution can improve the accuracy, efficiency, or effectiveness of the ML learning model for the computer vision task.”
