Animation modeling and media content description for authoring and presenting multimedia documents

The demo is based on VideoMadeus, an advanced environment for playing and editing multimedia documents. While tools such as GRiNS from Oratrix (http://www.oratrix.com/) or X-Smiles (http://www.xsmiles.org/) are based on the SMIL standard (http://www.w3c.org/AudioVideo/), ours uses the Madeus framework from the OPERA research project (http://opera.inrialpes.fr/). The Madeus model can be considered as an extension of SMIL, which provides the following features:

• enhanced separation of media content location, temporal information and spatial information;
• hierarchical, operator-based temporal model complemented with relations;
• rich spatial specification model with relative placements.

VideoMadeus is an implementation of the abstract animation model which is described in our article "An Abstract Animation Model for Integrating SMIL Basic Animation Elements with Multimedia Documents" to be presented at the conference ICME2002.

One of its main features is media fragment integration. It uses several views to display video and audio contents. These views allow the user to semi-automatically generate a media content description based on the MPEG-7 standard. This description is then used for specifying fine-grained synchronization between media objects. Using media content description in authoring multimedia documents brings such advantages as:

• tracking an object in a video (a video object for short),
• attaching hyperlinks to video objects (video objects are moving region),
• fine-grained synchronization: for example text can be synchronized with a video fragment (a scene, a shot or even an event).
• spatio-temporal synchronization: a text can follow a video object,
• extracting any part of a video/audio (even a video object) for integration with other media.

In addition, VideoMadeus provides a timeline view that is much more powerful than the usual flat timeline. Ours is hierarchical and supports editing of many temporal relations (meet, before, finish, during, equal, etc.). This is specially interesting to structure the video and audio media. It allows an author to easily locate the different parts of the media and to easily create temporal relations between media objects and fragments of the video/audio content.

This is a brief summary of VideoMadeus environment for authoring and presenting multimedia document. For more information please contact tien.tran_thuong@inrialpes.fr