���˴�Ƭ

The worry of unattended luggage at an airport terminal or train station might soon be a thing of the past, thanks to a new project that will see artificial intelligence (AI) used to identify when objects are left behind at busy public transport locations.

The ���˴�Ƭ’s Australian Institute for Machine Learning (AIML) has today signed an agreement for a six-month research project with to jointly explore how computer vision AI software may be able to detect objects using security camera footage and quickly determine if they might have been forgotten by passengers.

Computer vision is a field of AI that deals with how computers gain meaningful understanding of information from visual input such as video and images.

AIML’s researchers and Certis Group hope the technology might offer an improvement over requiring security staff to continuously monitor footage from multiple security cameras in large and crowded public transport stations, often for hours at a time.

Certis Security Australia’s Chief Executive, Kang Song Lim, understands the value of efficient, targeted technology in this space.

“We’ve listened to our customers, and we know they’re concerned about overcomplicated systems that don’t deliver the results they need. This partnership will create a market-first, a one-of-a-kind solution designed to detect any items that could pose a threat whilst reducing the work for our customers,” says Kang Song.

Certis Security Australia anticipates problems and provides innovative solutions across a wide range of security services, recognising that one size does not fit all. Backed by almost 100 years of rich history in physical security in Australia and more than 60 years of patented innovation in Singapore, Certis has grown from a security guard and escort unit to become a leading global integrated security services provider.

AIML researchers working on the project include Dr Ehsan Abbasnejad, Senior Future Making Fellow and Director of Human Centred Machine Learning at AIML; Dr Qi Wu, Senior Lecturer and ARC DECRA Fellow and Director of Vision and Language Methods at AIML; and Dr Yuankai Qi, AIML postdoctoral research fellow.

Dr Wu says the project will bring some interesting challenges, namely training AI software to detect the different types of items that passengers typically carry in public transport hubs.

“It’s not like traditional object detection where we have a predefined object,” Dr Wu says.

Dr Abbasnejad adds that detecting objects from video, rather than single images, brings further complexity.

“We want to detect that unknown object, but we also want to track it over time and recognise that it really is left behind,” he says.

Established in 2018 and located within South Australia’s Lot Fourteen innovation precinct, AIML is one of Australia’s largest machine learning research and innovation sites, with more than 160 members, and is ranked among world leaders in the application of AI, computer vision and machine learning to real world problems.