Current Press Releases

How can sensitive health data be used for research and innovation while protecting the individuals concerned? The Fraunhofer Institute for Digital Media Technology IDMT develops privacy-preserving technologies for this purpose. In the AVATAR-Transfer project, the institute and its partners from medicine, research, IT, industry and law are transferring these solutions into concrete applications by the end of 2028. At the annual congress held in Jena on October 7, 2026, around 60 participants from science, industry, medicine, politics and society discussed the privacy-preserving use of health data in practice. Fraunhofer IDMT presented two approaches in Jena: anonymization and digital watermarks are intended to protect EEG data. “Model Driven Privacy” incorporates data protection from the planning stage of data use.

Language skills constitute an important foundation for educational success, social participation, and personal development. Researchers at Fraunhofer IDMT are developing automated methods for analyzing children’s speech to identify developmental risks early on and support professionals in documentation, intervention planning, and progress monitoring. This technology addresses various application areas and can be integrated into digital diagnostic, educational, and therapeutic services.

Fraunhofer IDMT’s LEAP technology allows audio professionals to objectively measure and visualize dialog intelligibility, from monitoring media production and broadcast to quality control and post-production. To ensure consistent results across different audio solutions, leading manufacturers and the Institute’s Oldenburg Branch for Hearing, Speech and Audio Technology have now joined forces to establish a common LEAP implementation framework. At IBC 2026 in Amsterdam, visitors can discover current approaches by Fraunhofer IDMT and its partners for optimizing speech intelligibility in media content.

The Fraunhofer Institute for Digital Media Technology IDMT will present its expanded Content Verification Toolbox at IBC 2026, September 11–14 in Amsterdam. With a particular focus on audio and speech, the toolbox combines technologies for verification, forensics, provenance, tracking and content intelligence.

A new research project is set to advance automated wood sorting. In DatMuSSS – Data-Driven Multimodal Sorting of Logs in Sawmills – the German research institute Fraunhofer IDMT, Technische Universität Ilmenau and the German companies Vision & Control, PREMETEC Automation and Sägewerk Schwarzmühle are investigating how artificial intelligence can reliably assess the quality of roundwood and assign logs to the appropriate sorting category. Their goal is to develop a cost-effective inspection system that can be retrofitted to existing equipment, enabling small and medium-sized sawmills in particular to automate their sorting processes economically.

Drones can pose a security risk to airports, critical infrastructure, military facilities, and large-scale events. However, they are difficult to detect when they fly beyond visual line of sight, for example, when visibility is impaired by terrain, buildings, or weather. Fraunhofer IDMT in Oldenburg has developed an intelligent acoustic sensor solution that detects and localizes drones based on their characteristic sound patterns, even without visual contact. Fraunhofer IDMT experts will demonstrate the technology live on a demo vehicle during Drone Days from August 26 to 28.

Die Popakademie Baden-Württemberg untersucht gemeinsam mit dem Fraunhofer-Institut für Digitale Medientechnologie IDMT im zweijährigen Forschungsprojekt »Perspektive 2036 – Musikwirtschaft in der Zukunft«, wie generative Künstliche Intelligenz Musikproduktion, Distribution, Rechteklärung, Lizenzierung und Verwertung bis zum Jahr 2036 verändern kann. Ziel ist es, mögliche Zukunftsbilder der Musikwirtschaft und Musikkultur zu entwickeln und daraus konkrete Empfehlungen und Handlungsbedarfe für politische Entscheidungsträgerinnen und -träger abzuleiten. So sollen frühzeitig die Voraussetzungen für einen rechtssicheren, fairen und kulturell verantwortungsvollen Umgang von KI im Musikbereich geschaffen werden. Im Rahmen des Programms Insight II wird das Vorhaben vom Bundesministerium für Forschung, Technologie und Raumfahrt gefördert.In a two-year research project entitled ‘Perspective 2036 – The Future of the Music Industry’, the Popakademie Baden-Württemberg, in collaboration with the Fraunhofer Institute for Digital Media Technology IDMT, is investigating how generative artificial intelligence could transform music production, distribution, rights clearance, licensing and exploitation by the year 2036. The aim is to develop possible visions of the future of the music industry and music culture, and to derive concrete recommendations and areas requiring action for policymakers. The goal is to establish, at an early stage, the conditions for the legally sound, fair and culturally responsible use of AI in the music industry. The project is funded by the Federal Ministry of Research, Technology and Space as part of the Insight II programme.

New research project PADSE: Fraunhofer IDMT, Deutsche Welle, and Bauhaus University Weimar are developing novel methods for detecting audio deepfakes, using a person-centric approach for detecting synthetic and manipulated speech content.

During the AES Europe Convention in Copenhagen Professor Thomas Sporer was honored with the Fellowship Award from the Audio Engineering Society (AES). With this award, the AES recognized his outstanding contributions to perception-based audio quality evaluation as well as his long-standing commitment to the international standardization of audio engineering methods.

What does our brain reveal about us? Recording brain activity via electroencephalograms (EEGs) provides important insights for science and medicine. For example, modern AI systems can identify early signs of disease in these signals. However, the individual data patterns also allow conclusions about specific individuals. Experts refer to these as “brainprints”. The “NEMO” project, coordinated by Fraunhofer IDMT, has researched technologies to enhance data protection while simultaneously improving the availability of EEG data for scientific research. The researchers’ approach aims to significantly reduce the risk of reidentification in recorded EEG data, while preserving relevant information for data analysis. The research results have already been successfully demonstrated using sleep phase analysis as an example.