NUS Develops Algorithm to Thwart AI

A research team from the National University of Singapore (NUS) has developed a technique that it says will safeguard sensitive information in photos from AI-based techniques.

Privacy from AI

While the human eye can quickly scan through a few photographs at a time, computers can do far quicker. Today, machine learning (ML) algorithms used by social media platforms such as Facebook can quickly identify and tag users in uploaded photographs. Google Photos can also group uploaded personal photos to make it searchable by faces.

Dealing with modern threats to digital privacy hence necessitates finding a way to prevent machines from harvesting personal data from images.

The innovation from NUS is led by Professor Mohan Kankanhalli, the Dean of the School of Computing. His research team has developed a technique that makes changes to photos that are imperceptible to the human eye – but which render selected features undetectable by known algorithms.

The novel technique took six months to develop and relies on a study of 234 participants and a set of 860 images. By showing participants various images and having them identify those with visual distortion, the research team was able to identify factors that are picked up by humans.

Equipped with this “human sensitivity map”, the team fine-tuned their technique to apply visual distortion with minimal disruption to the image aesthetics – only injecting them into areas with low human sensitivity.

“It is too late to stop people from posting photos on social media in the interest of digital privacy. However, the reliance on AI is something we can target as the threat from human stalkers pales in comparison to the might of machines. Our solution enables the best of both worlds as users can still post their photos online safe from the prying eye of an algorithm.” said Prof Kankanhalli.

Users can use this technology to help mask vital attributes on their photos before posting them online and there is also the possibility of social media platforms integrating this into their system by default. The team says this will introduce an additional layer of privacy protection and peace of mind.

Photo credit: iStockphoto/bee32