AI Gives Cameras Brains

November 5, 0005

Computers are capable of thinking and learning thanks to AI, now cameras might be too. According to the Eurasia Review soon it will be possible that, “Cameras That Can Think.” Despite facial recognition software and other image based technology, an intelligent camera does not exist. The Universities of Manchester and Bristol, however, are on their way to designing an intelligent camera that will learn and understand what it sees.

Currently AI interprets data only after it has been recorded:

“This means AI systems perceive the world only after recording and transmitting visual information between sensors and processors. But many things that can be seen are often irrelevant for the task at hand, such as the detail of leaves on roadside trees as an autonomous car passes by. However, at the moment all this information is captured by sensors in meticulous detail and sent clogging the system with irrelevant data, consuming power and taking processing time. A different approach is necessary to enable efficient vision for intelligent machines.”

Bristol University researchers believe implementing Convolutional Neural Networks (CNNs), an AI that enables visual understanding, on the visual plane will classify information at thousands of frames per second. The CNNs do not need to record the images or process them. An AI camera could identify events or objects then send that information to a system, saving time, storage space, and being more secure.

An intelligent camera could soon be possible with the SCAMP architecture from Manchester University. SCAMP is a processor chip that is a Pixel Processor Array. A PPA has a processor embedded in each part of the array and every pixel communicates to each other to process a parallel form. This would be the necessary system to develop CNNs.

Cameras could become smarter tools than simply capturing images and video. Could AI cameras become tools for augmented reality own would they used as surveillance tools by bad actors? Probably both.

Whitney Grace, November 5, 2020

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta