Description
The thesis project develops and evaluates a vision-based system for detecting passenger falls and abnormal postures in autonomous buses. The work involves passenger detection, tracking, pose estimation, temporal analysis, robustness testing, and evaluation of accuracy, false positives and negatives, robustness, and processing latency. Candidates need programming experience in Python or C++, knowledge of computer vision and machine learning, a relevant technical background, and Swedish and English proficiency. The project is based in an office connected to the project, with most work expected to be done there.
