Raúl Díaz, Minhaeng Lee, Jochen Schubert, Charless Fowlkes
We are interested in algorithms that can find objects in images and video. Such algorithms have practical implications for applications such as image search and robotic navigation. Important and open issues include representation and learning - what makes a good model of an object, and how can one learn such a representation? To help answer such questions, we employ both generative and discriminative machine-learning techniques.