Proceedings of the 3rd Workshop on Human-Centric Music Information Retrieval 2025: Preface

Preface for the 3rd Workshop on Human-Centric Music Information Research (HCMIR25)⋆ Roser Batlle-Roca1,Ahyeon Choi2,Minhee Lee3,Lorenzo Porcaro4and Eunice Hong5 1Universitat Pompeu Fabra, Spain 2University of Iowa, USA 3Korea Advanced Institute of Science and Technology, Korea 4Sapienza University of Rome, Italy 5Seoul National University, Korea Abstract Music and technology have long intertwined, transforming how we create, share, and experience music. From composition to listening, advancements in Music Information Research (MIR) continue to reshape our relationship with music, enabling new forms of creativity and accessibility. However, alongside these innovations, critical questions arise about their broader implications. The 3rd edition of the HCMIR workshop seeks to explore MIR’s ethical, societal, and human-centred dimensions. How can we ensure MIR systems are inclusive, ethical, and aligned with human values? How do people perceive and engage with AI-driven music technologies? What role are artists playing in shaping these tools? How can design principles prioritize accessibility and usability while supporting human-AI collaboration in music creation and discovery? This workshop invites researchers, practitioners, and artists to engage in a multidisciplinary discussion on the societal impact, human experience, and future directions of MIR. By addressing these topics, we aim to foster a deeper understanding of how MIR systems can empower individuals, support creativity, and enhance the human experience of music. 1. Relevant Topics We invited the research community to submit short papers on the following relevant topics, but not limited to: 1. Ethics and Societal Impact of MIR • Ethical considerations in the development and deployment of MIR systems. • Societal challenges related to MIR technology, including environmental, political, legal, and social dimensions. • Implications of MIR systems on cultural diversity, bias, privacy, and inclusivity. • Long-term ethical and societal impacts of integrating MIR technology into everyday life. 2. Perception and Evaluation in MIR • User perception of AI-generated music compared to traditional compositions. • The role of AI identity (e.g., collaborative human-AI creation vs. fully autonomous generation) in shaping user experience. • Comparative studies on how people evaluate music created by different AI models. • Methods for assessing the aesthetic and emotional impact of AI-generated music in humancentered contexts. • Cross-cultural studies on music perception and AI-generated music. HCMIR25: 3rd Workshop on Human-Centric Music Information Research, September 20th, 2025, Daejeon, Korea ⋆ HCMIR25 is a satellite event of the 26th International Society for Music Information Retrieval Conference (ISMIR 2025), Daejeon (Korea). Envelope-Open[email protected] (R. Batlle-Roca); [email protected] (A. Choi); [email protected] (L. Porcaro); [email protected] (E. Hong) GLOBEhttps://roserbatlleroca.github.io/ (R. Batlle-Roca); https://lorenzoporcaro.me/ (L. Porcaro) Orcid0000-0003-3591-9378 (R. Batlle-Roca); 0000-0001-8304-599X (A. Choi); 0000-0003-0218-5187 (L. Porcaro); 0009-0007-1077-1377 (E. Hong) © 2025 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). 3. Human-AI Collaboration in MIR • Designing MIR systems that enhance creative collaboration between humans and AI. • The role of artists and musicians in shaping and improving MIR technologies. • The role of AI in traditional and novel music creative practices. • Tools, methodologies, and case studies that support co-creative processes between humans and AI. • Challenges and opportunities in fostering human-AI partnerships for music creation and curation. 4. User Experience (UX) and Interface (UI) Design in MIR • Best practices for designing accessible and enjoyable MIR systems. • Addressing diverse user needs, including those with limited technical expertise. • Evaluating the usability and effectiveness of MIR interfaces in music search, recommendation, and listening experiences. • Innovations in interface design to enhance engagement with MIR technologies. 5. Interdisciplinary Perspectives on Human-Centric MIR • Integrating insights from musicology, psychology, and human-computer interaction (HCI) into MIR research. • Cross-disciplinary approaches to addressing human-centric challenges in MIR development. • Collaborative frameworks involving diverse fields to improve the societal relevance of MIR systems. 6. Future Directions in Human-Centric MIR • Visionary approaches to embedding human values in MIR research and applications. • Emerging challenges and opportunities in aligning MIR systems with ethical and societal goals. • Strategies for ensuring MIR technologies evolve to prioritize human-centric values and creativity. 2. Accepted Contributions • ManyMusic: An open-access music audio dataset for human experiments on musical emotions (Seung-Goo Kim, Pablo Alonso-Jiménez, Dmitry Bogdanov and Daniela Sammler) • Shaping Structure Through Sound: Effects of Repeated Listening on Musical Segmentation in Musicians and Non-Musicians (Duri Bae, Ahyeon Choi and Kyogu Lee) • An Accessible Digital Musical Instrument for Inclusive Music Therapy (Raquel Lucena and Rafael Ramírez) • Examining Melodic Similarity Across Human, Computational, and Legal Perspectives (Roser Batlle-Roca, Lena M. Melo and Xavier Serra) • AImoclips: A Benchmark for Evaluating Emotion Conveyance in Text-to-Music Generation (Gyehun Go, Satbyul Han, Ahyeon Choi, Eunjin Choi, Juhan Nam and Jeong Mi Park) • Blaming the Algorithm: A Case Study on MusicVAE’s Bias Against Glitch-Modified POP909 Melodies (Weixi Zhai) 3. Keynote: Understanding the Human Experience of Music to Shape Future Technologies of MIR (Prof. Dr. Kyung Myun Lee) 3.1. Abstract As Music Information Retrieval (MIR) technologies evolve, their potential lies not only in analyzing acoustic features but in deeply understanding the human experience of music. In this talk, I will share work from our laboratory exploring how people perceive, enjoy, and create music, and how these insights can inspire the design of more human-centric MIR systems. Drawing on cognitive and physiological studies, we show that signals such as EEG and heart rate can reveal individual differences in music preference and expertise. We also extend this perspective to diverse populations, including individuals with hearing impairments and cochlear implants. While our work has so far focused on understanding how these listeners experience and emotionally respond to music, we see this as a first step toward imagining music generation and recommendation systems that adapt to their needs—making music more accessible and engaging for all. Finally, we investigate how shared music experiences can enhance social bonding and enjoyment, and discuss how these dynamics might be incorporated into MIR platforms. By weaving together cognitive, physiological, and social dimensions of music experience, I will open a discussion on how next-generation MIR technologies can put human values, personalization, and inclusivity at the center. 3.2. Biography Kyung Myun Lee is an Associate Professor at KAIST, where she directs the Music and Brain Lab and holds a joint appointment in the Graduate School of Culture Technology. She received her Ph.D. in Music Cognition from Northwestern University after earning degrees in Musicology and Psychology from Seoul National University. Her research focuses on the neural processing of music, interdisciplinary studies in music cognition and information retrieval, enhancing online concert experiences, and predicting music emotion and preference using biosignals such as EEG and EMG. She previously served as President of both the Asia-Pacific Society for the Cognitive Sciences of Music and the Korean Society for Music Perception and Cognition, and is currently Associate Editor of the journal Music Perception. 4. Program Committee • Alexandre D’Hooge (Algomus) • Arthur Flexer (Austrian Research Institute for Artificial Intelligence (OFAI)) • Blair Kaneshiro (Stanford University) • Bruce Ferwerda (Jönköping University) • Darius Afchar (Deezer Research) • Emilia Parada-Cabaleiro (Nuremberg University of Music) • Fabio Morreale (Sony AI) • Gerardo Roa Dabike (University of Salford) • Giovanni Gabbolini (Apple Music) • Helena Cuesta (Zoundream) • Ilaria Manco (Google DeepMind) • Karlijn Dinnissen (Utrecht University) • M. Stella Tavella (Musixmatch) • Magdalena Fuentes (New York University (NYU)) • Marius Miron (Earth Species Project) • Martín Rocamora (Universitat Pompeu Fabra) • Oriol Nieto (Adobe Research) • Pedro Sarmento (Moises / Music.AI) • Philip Tovstogan (AudioShake) Acknowledgments HCMIR25 is sponsored by the IA y Música: Cátedra en Inteligencia Artificial y Música (TSI-1009292023-1). Project funded by Secretaría de Estado de Digitalización e Inteligencia Artificial, y Unión Europea-Next Generation EU, and in collaboration between the Music Technology Group at Universitat Pompeu Fabra and BMAT. HCMIR25 is also supported by the project Algorithmic Auditing for Music Discoverability (AA4MD) which has received funding from the European Union’s Horizon Europe research and innovation programme under the Marie Skłodowska-Curie grant agreement No 101148443. More info at: https: //aa4md-project.eu/