Close your eyes and listen... surely you recognize each sound that surrounds you and you know exactly where it arises from. Certainly you can tell if there are cars driving through the street, when the coffee is ready or which colleague is talking to you and if he is standing behind or in front of you.

And now open your eyes, but caulk your ears... and you still exactly know what happens in your proximity. You can see the colleague speaking to you and you know that the coffee has been still being in progress.

The senses of hearing and seeing are complex processes that allow us not only to detect sounds and objects in our proximity, but also to identify and track them. We are able to follow a conversation even when in a noisy environment, or detect different sound events in a complex acoustic scene. We can recognize and understand persons and objects even if they partially disappear behind other persons or objects.

The opto-acoustic scene analysis has as its objective the detection and description of every acoustic and optic noticeable event in a specific scenario in order to achieve a robust environmental perception. The OPASCA (OPto-Acoustic SCene Analysis) research group investigates and develops methods and approaches which are targeted in building a system that can reproduce manlike perceptive behaviors.

Our main research area is the Collaborative Research Center 588 – Humanoid Robots, established by the Deutsche Forschungsgemeinschaft (DFG) and located at the Universität Karlsruhe since 2001. Further areas of applications are situated in the automotive industry and the surveillance domain.