معرفی
Mehrnaz Fani is an Adjunct Assistant Professor specializing in computer vision and sports analytics. Her research focuses on developing advanced algorithms for player tracking, rink localization, and event recognition in sports videos, particularly in ice hockey and soccer. She employs deep learning techniques and multi-task learning frameworks to enhance video analysis accuracy.
Her work emphasizes applications such as puck localization, jersey number recognition, and pose estimation of athletes. Fani's contributions bridge the gap between computer vision and sports science, enabling automated analysis of broadcast sports content.
Notable research trends include the use of neural networks (e.g., Hourglass, BendedNet) for geometric and pose estimation tasks, as well as parallel feature fusion for video structure analysis. Her methodologies often integrate temporal analysis and data augmentation to improve model robustness.
Despite no listed awards, her publications reflect a strong focus on practical sports video analysis challenges. She has advised no documented students and no grants are mentioned. Her work is part of a broader effort to automate sports data extraction and enhance fan engagement through advanced video processing.
