A QUT researcher has developed a visual recognition system for autonomous cars that mimics a human driver’s ability to recognise locations when approaching from a different direction and under radically different environmental conditions.
PhD researcher Sourav Garg, whose thesis is currently under examination, has been awarded first place in the SAGE Higher Degree Research Student Publication Prize for his article, Semantic-geometric visual place recognition: a new perspective for reconciling opposing views, published in The International Journal of Robotics Research.
Mr Garg, who is a researcher with the Australian Centre for Robotic Vision headquartered at QUT, said the prize was particularly pleasing to receive as the journal article was the culmination of his PhD research.
“Getting the award for this is very, very satisfying,” Mr Garg said.
Mr Garg was also one of the key researchers on the project with Professor Michael Milford to take an Artificial Intelligence (AI) system on a road-trip of south-east Queensland to ensure the autonomous cars of the future will be smart enough to handle tough Australian road conditions.
“Everything that moves, including vehicles and robots, needs to know where they’re located,” Professor Milford said.
“They usually do that by recognizing where they are against some map of the world they have in their head, in their computer.
“With cars, there are all sorts of challenges, like day turning into night and different types of weather. One of the particular challenges is obviously cars don’t always go the same direction.
“Sourav has been coming up with clever ways to solve this challenge.”
Mr Garg’s research outlines how navigation systems can take existing captured information and analyse it in a new way so that they can recognise locations they’ve previously gone past even though they are approaching it from a different direction, and landmarks are flipped and potentially partially obscured.
“As a human, we organically do these things – we don’t know exactly how we’re doing it. So one way is to interpret environment with human-like semantic scene understanding, enabled by deep machine learning ,” Mr Garg said.
“This is something which hasn’t been done before in research – coming from either direction and having different weather conditions.
“When I’m revisiting from the opposite direction, basically I’m trying to recognise and match ‘meaningful’ visual landmarks.”
Before undertaking his PhD at QUT, Sourav worked at a robotics research lab in India where he researched human and object tracking, including making a robot that could serve tea in an office environment.
The robot was programmed to carry a tray of teacups to workers, navigate an office space and avoid obstacles such as people crossing in its path.
SAGE Publishing has sponsored the SAGE Higher Degree Research Student Publication Prize at QUT since 2014.
Second place and $900 was awarded to Zhongtian Li for his paper, Corporate social responsibility employment narratives: a linguistic analysis, published in Accounting, Auditing & Accountability Journal. B.M.C. Randika Wimalasiri-Yapa was awarded third place and received $500 for her paper, Chikungunya virus in Asia-Pacific: a systematic review, published in the Open Access journal Emerging Microbes & Infections.
Rod Chester, QUT Media, 07 3138 9449, email@example.com
After hours: Rose Trapnell, 0407 585 901, firstname.lastname@example.org