Thien Luu Gia

Project: AI Camera for Traffic

Tasks: Low-Light Image Enhancement Algorithm for AI Camera; Video Coding with H265+.

Skills: Computer Vision, Image Processing.

Code Language: C++. Libraries: OpenCV, ffmpeg.

Achievement:

• Successfully developed a backlight compensation feature for traffic cameras, addressing the issue of AI camera license plate recognition in low-light or backlit conditions with over 90% accuracy.
• Applied the Retinex-Based Fast Algorithm for Low-Light Image Enhancement, in conjunction with OpenCV, C++, and Linux, to achieve a processing speed of 0.015s per image.

Project: Muscle Fatigue Detection using Machine Learning

• Developed a muscle fatigue prediction model with 90% accuracy using the Support Vector Machine algorithm.
• Research findings were presented at 3 international conferences:
• After-Fatigue Condition: A Novel Analysis Based on Surface EMG Signals | IEEE Conference Publication | IEEE Xplore
• Detection Of Muscles Fatigue Through Surface EMG Signals Utilizing Machine Learning Algorithm, BME10
• Nghi Tran Huu, Gia Thien Luu*, Bao Minh Pham, Philippe Ravier, Olivier Buttelli, Classification of Fatigued Electromyography Signals Using Support Vector Machine in Combination with Root Mean Square and Mean Absolute Value Features, Proceedings Of The 4th International Conference On Advanced Convergence Engineerin, Ho Chi Minh City, 2023.

Taught: Mechanics, Electromagnetism, Biophysics, Measurement in Medicine, Image Processing for undergraduate and graduate students.

Programming Language: MATLAB.

Skills: Teaching, Physics, Image Processing.

• Taught General Physics for undergraduate students.
• Conducted research in Biomedical Signal Processing.
• Supervised 5 graduate students.
• Taught statistics for telecommunications for Master's students.
• Led and managed the Physics Department.