M.Sc. Hanna Lukashevich is Head of the Semantic Media Technologies research group at the Fraunhofer Institute for Digital Media Technology IDMT in Ilmenau, Germany. She joined Fraunhofer IDMT in 2006 after receiving a Master's degree in Statistical Radio Physics from the Belarusian State University in Minsk, Belarus. Since 2014, she has led research activities in music and audio intelligence and currently heads a multidisciplinary team developing AI technologies for semantic media analysis.

For nearly two decades, her work has focused on Music Information Retrieval (MIR), automatic music analysis, audio signal processing, machine learning, and generative AI for music. She has led the development of technologies for music identification, similarity analysis, recommendation, metadata extraction, and large-scale semantic analysis of musical works. Her research aims to bridge computational methods, musicological knowledge, and practical industry applications, enabling machines to analyse musical structure, stylistic characteristics, content similarity, and inter-work relationships at scale.

More recently, her work has addressed the challenges that generative AI poses for the music ecosystem, particularly questions of transparency, provenance, attribution, and fair remuneration. Her research explores how methods from Music Information Retrieval, computational musicology, metadata analysis, and explainable AI can contribute to transparent and understandable attribution frameworks for generative AI. By analysing musical characteristics such as rhythm, harmony, melody, instrumentation, complexity, uniqueness, and similarity, these approaches aim to quantify relationships between musical works and support evidence-based mechanisms for attribution, provenance tracking, and value distribution that are understandable to creators, rights holders, and technology providers alike.

As a panelist on “From Training Data to Rights Holder: Closing the Attribution Gap,” she brings a technical and scientific perspective on how advances in music analysis, similarity assessment, metadata technologies, and explainable AI can support fair, transparent, and scalable attribution frameworks for the future music industry.