Neural networks have many practical applications and body scanning could be one of them
The US Department of Homeland Security (DHS) is launching a $1.5 million (£1.2m) competition to use neural network technology to improve airport screening procedures.
Working with Google, the agency is turning to data scientists to design computer algorithms that are able to automatically scan images captured by body scanners and spot hidden items at airport security checkpoints.
The contest will be run on Google-owned website Kaggle, a crowdsourcing platform specifically for machine learning and data analytics competitions.
According to Kaggle founder and CEO Anthony Goldbloom, the DHS contest will look to make use of neural networks, a technology that is rapidly increasing in importance in the world of machine learning and artificial intelligence.
Neural networks are essentially computer systems modelled on the human brain that are able to learn specific tasks by analysing huge quantities of data and are drawing significant investment from the world’s biggest technology companies.
For example, Google is using the technology to build AI systems that can create their own encryption and HPE recently partnered with Nvidia to power deep learning neural networks through graphics processing units (GPUs).
Neural networks have practical applications in a range of industries and DHS is hoping such a system can be used to more accurately read body scans to speed up the security process at airports and reduce the dependency on humans.
And, by crowdsourcing the competition on Kaggle, the agency will be able to tap into the skills of over a million data scientists to solve the extremely complex problem and transform airport security.
Airports are also upping their technology investments over here in the UK. London City is set to become the UK’s first airport to install a digital air traffic control tower, while Gatwick has installed 2,000 beacons to power an indoor navigation system based on augmented reality.