Audio Scene Analysis & Source Intelligence
- Acoustic scene understanding
- Multi-source tracking and regeneration
- Multichannel source separation
- Sound field reconstruction
Discuss a project
Discuss a project
LYKEN Labs
The intelligence layer for advanced audio systems. LYKEN Labs translates validated research into perceptually grounded, deployable solutions — from embedded devices to cloud-scale infrastructure.
From complexity to clarity
We combine core technologies, engineering, and a deep understanding of our clients' strategic context to translate validated research into tailored, perceptually grounded, and scalable solutions — seamlessly deployable across environments ranging from embedded systems to cloud-scale infrastructure.
Technologies
LYKEN Labs offers a Unified Audio Intelligence Layer that enables contextual understanding of audio scenes and provides a shared foundation for deployable, real-world solutions. Our integrated offering assembles signal processing, AI models trained with bespoke studio-grade datasets, and hardware platforms. Cloud solutions deliver state-of-the-art processing; edge solutions move intelligence closer to the user — on embedded systems, computers, and mobile devices.
We deliver fully operational, tailored solutions for
Audio / Video Equipment Manufacturers · Smartphone & tablet manufacturers · Chip DSP Companies · In-car Mobility
Telecom Operators · Broadcast / Streaming Services · Videoconferencing Providers
Audio / Video Plugin Editors · Videogame & Metaverse Developers · Venues · Music Education Institutions
Business users of the former market segments · Defence · Healthcare · Industrial R&D
Research
LYKEN Labs turns validated state-of-the-art audio research into a continuously evolving Audio Intelligence Layer — tailored to each client's use case, product architecture, and deployment environment.
Through REPERTORIUM's 60+ European research network and strong ties with INRIA and CNRS, our core technology layer stays continuously aligned with validated state-of-the-art audio innovations.
Client-specific by design, our solutions address deployment constraints from day one — from latency, compute, memory, and power to integration across embedded, edge, local, and cloud environments.
Each solution is validated through objective metrics, bespoke high-quality datasets, and critical listening by experienced sound engineers — ensuring it performs where it matters: in real-world acoustic and perceptual conditions.
About us
A unique combination of academic research, applied R&D, and perceptual audio expertise — "musical ears" — translating scientific advances into industrial-grade, perceptually validated solutions.
Access to premium datasets via professional studios (Studio Sequenza, The Spheres) — enabling perfectly tailored training and evaluation data, a key competitive advantage in audio AI.
Direct access to cutting-edge research through a network of 60+ researchers (REPERTORIUM, €3M European HORIZON R&D project), keeping our roadmap aligned with the state of the art.
LYKEN Labs was founded following the successful completion of REPERTORIUM, which developed sound source separation, physics-informed spatial audio reconstruction, and audio-to-score alignment solutions for classical symphonic music — the most complex and unforgiving of all sound fields. Engineered for symphonic complexity, our technologies scale confidently across a broad range of real-world audio applications.
CEO
Entrepreneur. Executive management, consulting, R&D, product development.
École Polytechnique (X) · Télécom Paris
CINO
Innovation strategy, product vision, research coordination. Co-author & coordinator of REPERTORIUM.
University of Oxford
Contact us
Tell us about your system, your constraints, and your ambition. We'll get back to you with a thoughtful first read — usually within two working days.