A smartphone-based system could teach machines to see and accelerate development of driverless cars.
Two technologies designed by researchers from the University of Cambridge, which use deep learning techniques to help machines to see and recognize their location and surroundings, could be used for the development of driverless cars and autonomous robotics – and can be used on a regular camera or smartphone.
The separate but complementary systems are freely available online. Although the systems cannot currently control a driverless car, the ability to make a machine ‘see’ and accurately identify where it is and what it’s looking at is a vital part of developing autonomous vehicles and robotics.
Alex Kendall, a PhD student in the Department of Engineering, said:
“It’s remarkably good at recognising things in an image, because it’s had so much practice. However, there are a million knobs that we can turn to fine-tune the system so that it keeps getting better.”
Professor Roberto Cipolla, who led the research, said:
“Vision is our most powerful sense and driverless cars will also need to see. But teaching a machine to see is far more difficult than it sounds.”
source University of Cambridge