Fraunhofer Institute for Digital Media Technology IDMT

Director: Prof. Dr.-Ing. M.S./SUNY Joachim Bös

Contact: Dr. rer. nat. Jens-E. Appell

The Fraunhofer Institute for Digital Media Technology IDMT conducts application-oriented research in the field of audio-visual technologies at its locations in Ilmenau and Oldenburg. The scientists develop innovative solutions in the field of sound reproduction and comfort, safety and security, assistance as well as automated production.

The Fraunhofer IDMT has many years of know-how in design of 3D sound solutions and planar speaker systems. Seat-specific solutions are increasingly in demand for sound reproduction in the passenger compartment – ranging from volume adjustment to individual hearing zones. The “Personal Sound Zones” technology developed at the Fraunhofer IDMT allows different media content to be played at different seats in the vehicle.

Using computer-based models of human auditory perception, it is possible to predict the intelligibility of infotainment systems or the audibility of warning signals under the influence of background noise. In order to not just evaluate but also optimise speech intelligibility, the Fraunhofer IDMT offers signal processing technologies which analyse the current acoustic ambient conditions and adapt the audio signal in real time. Integration of methods from hearing aid technology allows audio reproduction to also be optimised for persons with impaired hearing.

In vehicle engineering, it is already possible to determine at the development stage how variations in the product characteristics will affect the perceived sound quality of operating noises. The Fraunhofer scientists develop customer-specific evaluation models and software solutions which provide meaningful results over and beyond the information provided by the usual characteristic values.

The scientists also work on speech recognition systems which function robustly even under conditions with interfering noise and at a greater distance from the microphone. The recognisers are therefore suitable both for use in the vehicle interior as well as for operating functions outside and in tough conditions. Further applications for computer-based speech recognition include documentation, safety and training applications in the traffic and transport field, e.g. for maritime communication. The technology can be integrated in embedded systems and can be used without an internet connection.

In addition to speech, computer-based detection is also possible for other acoustic events – such as signals relevant to safety and security or operating noise irregularities. Possible applications include smart vehicles, traffic monitoring in smart cities as well as acoustic quality control and status monitoring in Industry 4.0 scenarios. The portfolio extends from optimised signal recording with microphone arrays and signal pre-processing through to recogniser technologies in embedded systems and sensor networks.

In order to avoid accidents, the gaze direction of the driver is continuously monitored by means of eye tracking technology to provide an appropriate warning in case of driver fatigue or lack of attention. Operation of other assistance systems, such as automatic lane keeping or brake assist, can be coupled to this. Thanks to video-based monitoring of the passenger compartment, it is also possible to determine the position of the vehicle occupants – a life-saving measure when it comes to deciding which airbags should be deployed in an accident and which not.