PETALING JAYA: Team DoubleY from Universiti Tunku Abdul Rahman (UTAR) has made waves in the world of data science by clinching first place in the prestigious IEEE BigData Cup 2024.
The dynamic duo behind the victory are Wong Yi Jie, 26, a Year Two PhD (Engineering) student from UTAR’s Lee Kong Chian Faculty of Engineering and Science, and UTAR alumnus Khor Yin Loon, 24, who currently works at Intel Corp.
Their groundbreaking solution leveraged satellite imagery, generative artificial intelligence (AI), and deep learning to improve building footprint extraction algorithms.
This innovation not only sets a new benchmark in data science but also promises transformative applications in urban planning, disaster response and environmental monitoring.
The competition, which ran from May to September 2024, was fiercely contested, drawing 263 submissions from 68 global teams.
“The most memorable moment was when we finally ranked first place. We didn’t skyrocket to the top. Instead, we climbed the leaderboard steadily through various stages,” Wong said in an interview yesterday.
He said the team’s winning solution involved utilising deep learning and designing an AI model to extract and identify all buildings from satellite images.
To ensure robustness, they developed an image generator to produce synthetic images, enhancing the training dataset and enabling their AI model to perform well even with varied test data.
Wong acknowledged the hurdles they faced, particularly limited resources for experiments.
“Fortunately, we were able to come up with creative solutions to overcome this challenge,” he said, adding that their testing process, capped at five daily submissions, required careful planning.
The implications of their work extend beyond the competition as Wong explained that the AI model could assist in disaster response by identifying affected areas and optimising relief efforts.
“In regions undergoing rapid urbanisation, our solution can assist in mapping and infrastructure planning,” he said.
Looking ahead, the team aims to delve deeper into geospatial AI and explore its applications in climate change analysis and renewable energy planning.
“Specifically, we will participate in the upcoming 2025 Ernst and Young Open Science AI and Data Challenge, which focuses on building predictive models to detect urban heat events.
“I strongly encourage anyone interested in AI and data science to join,” said Wong.
In a statement, UTAR praised the team’s achievement, noting that Wong and Khor presented their winning solution virtually.
“The team’s outstanding performance not only earned them US$1,000 (RM4,504), a trophy and certificates but also established UTAR as a leader in innovative data science solutions on an international platform,” it said.
IEEE is the Institute of Electrical and Electronics Engineers.
