Congratulations to the team from the AI Club - Faculty of Computer Science at UIT for securing the TOP 4 position in the problem branch of Road Object Detection in Fish-Eye Cameras of the 2024 AI CITY CHALLENGE, held within the framework of the CVPR 2024 conference - the most prestigious conference in the field of computer vision. The problems in the 2024 AI CITY CHALLENGE aim to utilize AI in areas such as smart transportation and retail. The competition attracted numerous research teams from universities, research institutes, and businesses worldwide.
Road Object Detection in Fish-Eye Cameras is a problem of detecting objects from fish-eye cameras (images often distorted, creating a fish-eye effect) - known for their ultra-wide-angle lenses, capable of providing video surveillance with dynamic viewpoints, including a 180-degree and 360-degree panorama. These cameras help reduce blind spots with the ability to produce multi-directional viewpoints to increase coverage without the need for additional cameras. The problem is set in the context of smart transportation, so teams need to identify objects including buses, bicycles, cars, pedestrians, and trucks. The main challenges that teams face include: distorted object images, small objects, and imbalanced data.
The team from the AI Club consists of first-year and second-year students from the Faculty of Computer Science at UIT, University of Information Technology - National University, Ho Chi Minh City, who have made efforts to experiment with various solutions and secured the Top 4 position on the competition scoreboard. This is a commendable effort by UIT students within the competition, which featured strong teams from businesses, institutes, and schools worldwide.
Once again, congratulations to all of you.
For more details, please visit: https://web.facebook.com/uit.cs.aiclub/posts/pfbid02csnFYqw36DJgUu2psEVK58SNa6Cm7R6dwHvmjSwGd9XAPMSWdzjAgrDUVx2jq7Dol
Written by: Dong Xanh
Translated by: Ngoc Diem