About this role
Summary: Develops speech AI algorithms for hearing aids, covering architecture exploration, model training, evaluation and deployment on resource-constrained wearable devices.
If you’ve worn a pair of glasses, we’ve already met.
We are a global leader in the design, manufacture, and distribution of ophthalmic lenses, frames, and sunglasses. We offer our industry stakeholders in over 150 countries access to a global platform of high-quality vision care products (such as the Essilor brand, with Varilux, Crizal, Eyezen, Stellest and Transitions), iconic brands that consumers love (such as Ray-Ban, Oakley, Persol, Oliver Peoples, Vogue Eyewear and Costa), as well as a network that offers consumers high-quality vision care and best-in-class shopping experiences (such as Sunglass Hut, LensCrafters, Salmoiraghi & Viganò and the GrandVision network), and leading e-commerce platforms.
Join our global community of over 190,000 dedicated employees around the world in driving the transformation of the eyewear and eyecare industry.
Discover more by following us on LinkedIn!
Your #FutureInSight with EssilorLuxottica
Are you willing to pioneer new frontiers, foster inclusivity and collaboration, embrace agility, ignite passion, and make a positive impact on the world? Join us in redefining the boundaries of what’s possible!
Based in Sophia Antipolis, France, Pulse Audition is a fresh start-up like team within EssilorLuxottica. The team consists of experts in AI, Audio Signal processing, Audiology, Hardware and Software who are working together to developp cutting-edge technologies to be integrated into smart eyewear products, to improve the social inclusion of people suffering from hearing loss.
Your role
We are looking for an AI/ML Speech Algorithm Engineer who will join our Audio AI team to work on cutting-edge speech AI technologies for hearables, contributing across the full research-to-productization pipeline — from exploring novel architectures and proof-of-concept work through productization and deployment on resource-constrained wearable devices.
Main responsibilities:
- Contribute to speech AI algorithms end to end — speech enhancement, acoustic feedback cancellation, beamforming, acoustic scene recognition — from architecture exploration through to a verified feature on a shipping device.
- Design, train and evaluate deep learning architectures against hard real-time latency, MIPS and footprint budgets, incorporating quantization, pruning and other model optimization techniques for deployment.
- Build out the evaluation stack: objective and perceptual metrics, listening tests with audiology colleagues, and regression testing to verify model improvements.
- Generate and curate the data the models need — noise corpora, room impulse responses and acoustic simulation, multi-microphone and binaural recordings.
- Contribute reusable training and testing infrastructure across tasks.
Main requirements:
- MSc or PhD in Electrical Engineering, Computer Science or a related STEM field, with real depth in deep learning and signal processing (a BSc with equivalent demonstrated depth also works).
- 5+ years building audio or speech ML algorithms, including at least one algorithm personally carried from research into a shipped product.
- Hands-on model development and production experience in Python, with PyTorch or TensorFlow.
- Track record of deploying models to real-time or embedded targets under latency, memory and compute constraints, iterating on the basis of measurements rather than intuition.
- (Preferred) Hearing-aid-class signal processing, or experience with hearables and wearable audio products.
- (Preferred) End-to-end audio system understanding: transducers, microphones, audio ICs, acoustics, embedded DSPs and NPUs.
- (Preferred) Familiarity with model optimization techniques — quantization-aware training, pruning, knowledge distillation, and other methods for compressing models without sacrificing performance.
- (Preferred) Cloud-based training environments (GCP or equivalent) and experience with GPU cluster workflows.
- (Preferred) Exposure to regulated or medical-device product development.
- (Preferred) Publications or patents in speech, audio or applied deep learning.
FR: Cette offre est publiée en anglais en raison du contexte international du poste et de la nécessité d'utiliser l'anglais dans l'exercice des fonctions.
EN: This job posting is published in English due to the international nature of the role and the requirement to use English in the day-to-day performance of the position.
Our Diversity, Equity and Inclusion commitment
We are committed to creating an inclusive environment for all employees. We celebrate diversity and provide equal opportunities to all, regardless of race, gender, ethnicity, religion, disability, sexual orientation, or any other characteristic that makes us unique.
#LI-BD1