The proposed research is oriented to introducing new applications of commercial interest in audio processing. The specific goals are:

1. Sound source separation and localization in real time based on signal models.
2. Score-to-audio alignment for virtual accompaniment.
3. Distributed processing of sound information.
4. Global decision taking from distributed sound information.
5. Development of applications in real environments for entertainment.

Consequently, this Project aims to apply recent research results in sound applications with comercial interests. The designed applications will interact with the user in an active way. For this reason, audio processing must be performed in real time in order to obtain devices that can interact with the users.

Some of our applications will be devoted to professional musicians and audio listeners. The key technology to implement these applications will be score alignment. This technology will be controlled by an expert system that will be able to follow in a dynamic way the tempo variations of a music performance. Using this tool, we aim to develop a set of applications with commercial interests as virtual accompaniment, technologic concerts or "listener" score. In the last one, for example, the system provides information related to the measure playing at the current instant and, also, the user can select an enhancement of the instrument that he/she wants to follow.

Also, we pretend to develop some applications for sound masking and cancellation. For example, we want to develop an active cancellation of speech in cinema in order to cancel the broadcast language and substitute it by another language. Here, we will make use of the results in signal separation.