MathWorks, a leading developer of technical computing software, announced the winners of the MATLAB AI Model Challenge on the 27th.

The challenge ended on January 31, 2025, and participants used MATLAB and Simulink to model ideas that could revolutionize real-world or industrial processes.

A panel of judges, comprised of MathWorks engineers, evaluated code submitted by entrants via the MATLAB user community, File Exchange.

This challenge had four award categories: the MathWorks Award, the Popularity Award, the Rating Award, and the Prolific Award. The winners were selected based on the following criteria: relevance to the topic, technical excellence, usability of MATLAB and other MathWorks tools, number of downloads, and star ratings.

The MathWorks Award went to Mita Nurhayati of Kyungpook National University for her project, “Explainable SHAP-based Neural Network Regression Model.” The project was highly praised for directly implementing the SHAP algorithm and visualizing the results.

The Popularity Award went to Samsung Electronics' Geunsik Kim for his project 'USB Camera-based Real-time Deep Learning Emotion Recognition System'. This project introduces a deep learning development workflow in a MATLAB environment and implements real-time data processing.

The award went to Youngsoo Choi of the National Institute of Scientific Investigation for his project titled ‘Deep Learning System for Vehicle Tracking and Wheel Detection.’ This project used the latest AI model called SiamFC to predict the speed of a vehicle.

The multi-work award went to the Korea Research Institute of Bioscience and Biotechnology team (Jo Su-bok, Kim Jong-hwan, Jeon Jong-beom, and Lee Dong-yoon). The team utilized the massive data of the 1000G Project, a public genome data, and provided access to the dataset through GitLarge file storage.

“Many engineers around the world are sharing innovative codes through the MATLAB community and increasing the efficiency of AI system development,” said Jongmin Lee, CEO of MathWorks Korea. “Through this challenge, we hope that more developers will be able to realize innovative ideas and grow by utilizing free open source code developed with MATLAB and Simulink.”

The winning entries for this challenge can be found on MathWorks' File Exchange (https://kr.mathworks.com/matlabcentral/fileexchange) and are available for free download.