About
Adrien Lefevre
Research engineer in machine learning, based in Lausanne.
Research
I am a research engineer at EPFL’s VITA Laboratory, working on machine learning for audio restoration. My work concerns generative models for degraded music recordings and the experimental methods used to evaluate their outputs.
I also contribute to OpenRestor, an open-source benchmark for musical audio restoration. This work involves the infrastructure needed to prepare data, construct degradation experiments and compare restoration systems.
Background
I trained in engineering at ESILV and specialised in Creative Technology at the De Vinci Innovation Center / Institute for Future Technologies. My Master of Science thesis examined tangible and augmented interfaces for music practice and sign-language learning.
Before my current work at EPFL, I worked as a Visiting Research Assistant at MIT Media Lab and as a Research Assistant in Haptics at Saarland University’s Human Computer Interaction Lab. My projects have combined computer vision, electronics, interactive systems and user-oriented research.
I am a co-author of ReMirrorFugue, published at CHI 2025, which examines the experience of presence through recorded piano performances. My work on an interactive paper score was published as a demo at NIME 2022.
Music and other interests
I studied piano, music theory and singing at the Conservatory and Music School of Boulogne-Billancourt. This musical background has informed several of my projects in audio and interaction.
Outside research, I enjoy photography and video editing, visual effects, design, kayaking, martial arts, table tennis and bouldering.
Contact
For research collaborations and professional enquiries:
Curriculum vitae ↗