Skip to main content

AI security cameras inspired by frog eyes could be coming soon

(Image credit: Pixabay)

University researchers are exploring ways to make security cameras more intelligent.  The new approach would use artificial intelligence to create more efficient camera systems that only record high-level, relevant events taking place.

The research is part of a collaboration between the Universities of Manchester and Bristol that aims to improve existing methods employed by camera systems to process information. 

Currently, visual sensors record everything before sending it to GPUs for processing, causing systems to become clogged with huge amounts of irrelevant data.

“To create efficient perceptual systems we need to push the boundaries beyond the ways we have been following so far,” Walterio Mayol-Cuevas, Professor in Robotics, Computer Vision and Mobile Systems at the University of Bristol, explained. “We can borrow inspiration from the way natural systems process the visual world — we do not perceive everything — our eyes and our brains work together to make sense of the world and in some cases, the eyes themselves do processing to help the brain reduce what is not relevant.”

The future of A-Eye

The two universities are exploring the use of Convolutional Neural Networks, a form of AI algorithm, to classifying thousands of visual frames per second, without having to record any of the images. The idea takes inspiration from a frog’s eye, which detects fly-like objects directly.

If the technology can be developed further, it could lead to a complete revamp of the security space, ushering in a future where AI cameras integrate the sensing, processing, and storage of visual data at the pixel level.

As well as taking cues from the natural world, the AI cameras also rely heavily on a new camera-processor chip called SCAMP. The chip has a processor embedded in every pixel, making it ideal for AI algorithms looking to rapidly assess visual data.

Via SciTech Daily