Researchers from the University of Central Florida have unveiled new artificial intelligence that can see, recognise shapes and identify objects. This technology could be used in robotics or to improve autonomous car systems.
Could sight be a sense soon to be unlocked by artificial intelligence? This is the basis of a project led by researchers at the University of Central Florida.
Thanks to a device capable of reproducing the retina of a human eye, their creation could lead to new, more powerful AI systems with new capabilities.
In practical terms, the technology could allow AI to instantly understand what it’s actually looking at. The most obvious applications for such technology are autonomous cars and robotics.
Another benefit of this technology – outlined in a recent study, published in the journal, ACS Nano – is that it is more powerful than the eye in terms of wavelength range. That is, it can perceive ultraviolet as well as visible light. For self-driving vehicles, the device’s versatility could offer safer driving in a range of conditions, as Molla Manjurul Islam, the study’s lead author, explains.
“If you are in your autonomous vehicle at night and the imaging system of the car operates only at a particular wavelength, say the visible wavelength, it will not see what is in front of it,” explains Molla Islam. “But in our case, with our device, it can actually see in the entire condition.”
According to the researchers, there is currently no device of this type capable of operating equally well across ultraviolet, visible and even infrared wavelengths. These capabilities give the system a unique character. The technology is also highly compact.
The intelligent imaging technologies currently available function in several separate stages, from the sensing to the memorisation and processing of data. “We had devices, which behaved like the synapses of the human brain, but still, we were not feeding them the image directly,” says Tania Roy, an assistant professor at the University of Central Florida.
“Now, by adding image sensing ability to them, we have synapse-like devices that act like ‘smart pixels’ in a camera by sensing, processing and recognising images simultaneously.”
For now, the accuracy rate is around 70% to 80%, and this should continue to improve as the researchers continue to develop the system. The scientists estimate that the technology could be ready in the next five years.