Bilkent University
Department of Computer Engineering


Prototypes: Exemplar based video representation


Özge Yalçınkaya
MS Student
(Supervisor: Assoc. Prof. Dr. Selim Aksoy)
Computer Engineering Department
Bilkent University

Action recognition is a widely studied topic and many works have introduced over the years. However, video classification with weakly-labeled large scale web data continues to be a challenge due to their noisy content despite recent improvements. By motivating the success of discriminative methods which are suggested for image classification, we introduce a novel video representation that is based on selected distinctive exemplars. We call these discriminative exemplars as "prototypes" which are chosen from each action class separately to be representative for the class of interest. Then, we used them to describe the entire dataset. Following the traditional supervised classification methods and utilizing the available state-of-the-art low-level features, we show that even with simple selection and representation methods, use of prototypes can increase the recognition performance. Moreover, by reducing the training data to the selected prototypes only, we show that less number of carefully selected examples could achieve the performance of a larger training data. In addition to prototypes, we adapted the previously proposed irrelevant data elimination method "AME" to action recognition and give the experimental results which are comparable to or better than the state-of-the-art studies on benchmark video datasets UCF101 and ActivityNet by exploring the effects of its parameters with both low and deep-level features.


DATE: 27 June, 2016, Monday @ 14:00