Mohamed E Hussein

My main research interest is computer vision. My computer vision research spans two subareas. The first subarea is human detection in still images. The second one is vision computing on Graphics Processing Units (GPUs). My research in human detection covered several aspects of the problem. Particulary, it covered feature extraction, classifier training, evaluation of detection algorithms, and human detection system design and integration. My research in vision computing on GPUs included development of efficient implementations for Graph Cut and Affinity Propagation on GPUs. From my experience in these two subareas, I realized the power of learning from large datasets, and the power of parallel computing. This inspires my current research focus, which is leveraging the power of parallel computing to enable large scale learning for computer vision applications.

More details to come ...