BrainStorm'19 - Top Ten

Oct 8, 2023   | #Deep Learning  #Computer Vision  

Competition Title	Brainstorm 2019, All island Competition
Organizer	Biomedical Society, University of Moratuwa
Year	Dec 2019
Place	Top Ten
Team	Nuwan
Code	Paper

Description

"Brainstorm" was an exciting journey into the world of innovative solution development. We took part in a cutting-edge competition organized by the UoM Bio-medical Faculty, where our goal was to enhance EEG medical imagery using state-of-the-art deep learning techniques. To achieve this, we built a novel solution based on image super-resolution concepts, implementing a robust model using the fully convolutional UNet architecture. Our model exhibited exceptional performance despite limited training data. Furthermore, we developed a user-friendly desktop application that allowed for interactive EEG image enhancement.

Technologies

Methodologies

Image Super-Resolution Fully-Convolution Neural Network Desktop Application Development

Key Achievements

• Innovative Solution: Brainstorm offered a novel approach to enhance EEG medical imagery, addressing a critical need in the healthcare industry.
• Deep Learning Expertise: Leveraging deep learning and image super-resolution techniques, we achieved impressive results, even with limited data.
• User-Friendly Application: Our desktop application made our solution accessible to a broader audience, bridging the gap between advanced technology and user-friendliness.