Video Shot Representation Using a Neural Network
Nabil Madrane, Said Jai Andaloussi, Abderahim Sekkaki · 2015
This paper describes a work concerning the use of neural networks for learning both the static and dynamic structure of a video clip and thus generating signatures suitable for indexing and retrieval. We show how to design a neural network suitable for representing a video shot and we describe how the inner weight data of the neural network can be used for video shot retrieval and extraction of signatures. This paper also presents preliminary experimental results of our approach.