The main goal of the Intelligent Browsing in Multimedia Content is demonstrate the capacity of audio indexing to facilitate access to multimedia content, and provide a web tool allowing the user to access his favorite multimedia contents anywhere and at any time.
This demonstrator has 2 main attention focuses:
On the generation of football match summaries, no human interaction is needed. The system records the football match and automatically extracts the highlight moments. And depending on the user preferences generates longer or shorter summaries.
The final interface will be multimodal, the user will be able to navigate using the mouse and the keyword like any other web, but they will be also able to easily navigate through the summaries with their voice. They could select the summaries they want to see and easily navigate through it, skipping or repeating different highlight events detected by the system.
Although the demonstrator includes both systems: the football match summary and the human-machine interaction to watch those summaries, from now on we will describe the HMI part, since this one is the one the user will receive. Although the football summary is essential to the overall demonstrator, it works on the server part and it is not necessary to install it to the users.
On the figure the demonstrator architecture is shown.
Interface hub deals with petitions from the speech recognizer and the keyboard, and it manages the web application, and the video player.