Gerhard Widmer

Gerhard Widmer is a Full Professor at the Institute of Computational Perception within the Faculty of Engineering and Natural Sciences at Johannes Kepler University Linz (JKU). He leads research at the LIT Artificial Intelligence Lab and serves as principal investigator for major projects including the FWF-funded Cluster of Excellence "Bilateral Artificial Intelligence" (2024-2029) and the EU project "Whither Music?" (2022-2026), focusing on AI applications in music cognition and performance analysis.

His research spans Music Information Retrieval (MIR) with emphasis on computational modeling of musical expressivity, audio processing, and machine learning applications. Key areas include sound event detection, music transcription, language-based audio retrieval, and modeling of expressive musical performances using deep learning techniques. His work bridges artificial intelligence with musicology to develop systems capable of understanding and generating human-like musical expressions.

Analysis of recent publications (2024-2025) reveals a strong trend toward diffusion models for audio generation/manipulation, robustness studies in music emotion recognition, and efficient architectures for sound event detection. There is significant focus on multimodal approaches integrating audio with textual descriptions, knowledge distillation for model compression, and real-time processing solutions for symbolic music analysis and score following.

No specific scientific awards are documented in the provided materials.

Professor Widmer has supervised 8 research works and manages multiple funded projects:

• Cluster of Excellence "Bilateral Artificial Intelligence" (FWF, 2024-2029): Investigating human-AI collaborative intelligence
• "Whither Music?" (EU, 2022-2026): Exploring AI's impact on musical creativity
• CoMIRVA: Developing music information retrieval visualization tools
• Sony-CSL project (2021-2024): Industry collaboration on music cognition

He directs the Institute of Computational Perception at JKU and is central to the LIT Artificial Intelligence Lab, maintaining active collaborations with the International Society for Music Information Retrieval (ISMIR) community through conferences, shared datasets, and open-source tools like Partitura for symbolic music processing.