Projects Physics and Perception of Environments

Date of the last update : 10/08/2020

Ongoing projects of the "Physics and Perception of Environments" team.

PhotoLED : Future photometry based on solid-state lighting products

Developing LED-based standard lamps for solid-state light calibration

Solid-state lighting, which uses light-emitting diodes (LEDs), is globally replacing traditional incandescent lighting, due to lower power consumption and greater durability. Photometers are used to measure the performance of lights, and are calibrated using standard lamps to ensure the accuracy and consistency of measurements. However, the standard lamps used for calibration are currently based on incandescent lights, not LEDs. This project will develop new standard lamps based on LEDs and new measurement techniques for defining the properties of solid-state lights. The results will be used by National Measurement Institutes and test laboratories to accurately calibrate solid-state light photometers and will give European industry an advantage in the development of new commercial standard lamps. These outputs will result in a more reliable classification of the energy efficiency of solid-state lighting, increasing consumer confidence in this new greener technology.

Participating EURAMET NMIs and DIs

Other Participants

  • Danmarks Tekniske Universitet (Denmark)
  • ENTPE (France) - Sophie JOST
  • LMT Lichtmesstechnik GmbH Berlin (Germany)
  • OSRAM GmbH (Germany)
  • OSRAM Opto Semiconductors GmbH (Germany)
  • Philips Lighting B.V. (Netherlands)

Speech2Ears: A predictive model to support hearing-aid processing restoring spatial perception for speech recognition in realistic environments

The 3-year project Speech2Ears is funded by the Fondation Pour l’Audition. It is coordinated by Mathieu Lavandier (ENTPE), in collaboration with Jörg Buchholz (Macquarie University, Australia) and Virginia Best (Boston University, USA). The project will involve the PhD projects of Thibault Vicente and Luna Prud’homme.

In our rapidly ageing population, hearing problems are continuously increasing. When listening in a noisy environment, the ability to spatially separate a target voice from competing sound sources can greatly improve intelligibility. However, this ability is generally reliant on having two normally functioning ears — hearing loss and current hearing-aid processing can greatly reduce this benefit. Providing hearing-impaired people with appropriate hearing devices is crucial for them to be able to take part actively in the society. The aim of Speech2Ears is to facilitate new hearing-aid processing strategies, which could restore this spatial benefit, through the development of a binaural speech intelligibility prediction model that can account for individual hearing loss and hearing aid processing in complex, realistic scenarios, including reverberation and competing speech sources. Such a model is required for a relevant assessment of the performances of hearing aids, thus crucially conditioning their development. It will also be very useful for the improvement of building comfort and accessibility, which currently rely on predictions assuming normal-hearing capabilities.
The project will rely on international collaborations between three highly complementary research institutions: ENTPE/University of Lyon (France), Macquarie University/NAL (Australia) and Boston University (USA). Speech2Ears aims to develop in two complementary directions a normal-hearing model proposed by ENTPE to predict intelligibility against noise and reverberation. The first part will develop the model by taking into account the influences of hearing impairment and hearing aids (PhD of Thibault Vicente). The second part will aim at predicting intelligibility in the presence of speech maskers as compared to noise (PhD of Luna Prud’homme).  We will then be able to infer the amount of interference caused by competing talkers that disrupts intelligibility even when the target is sufficiently audible, and the potential influence of hearing impairment and hearing aids on this interference.
ENTPE developed a binaural intelligibility model for normal-hearing listeners, Macquarie University developed an experimental set-up to measure the performance of hearing-impaired listeners in both simple and more realistic listening situations, and Boston University developed methods specifically designed for studying attentional factors influencing performance in speech mixtures. Creating a synergy between these complementary partners, Speech2Ears will produce knowledge and tools supporting hearing-aid developments and benefiting the public.

Mathieu Lavandier
funded by the Fondation pour l’audition

Partners :
Jörg Buchholz
Virginia Best

PhD Students:
Thibault Vicente (ENTPE)
Luna Prud’homme (ENTPE)



Rôle de la naturalité sur les interactions VIsuelles et SONores en environnement urbain extérieur et souterrain
L’amélioration du bien-être en ville est un enjeu majeur de la gestion et de la planification urbaine. La renaturation des villes est l’une des stratégies très largement mobilisée pour y parvenir. Nous proposons d’étudier l’impact de la naturalité des espaces publics urbains sur les perceptions visuelle et sonore. Deux contextes urbains seront étudiés dans le cadre de ce projet de recherche : (1) Des espaces publics souterrains (quais de métro) ; (2) Des espaces publics extérieurs (espaces verts aux abords des voies de communication). Ils ont en commun de présenter d’importants bruits de trafic qui peuvent nuire au bien-être des citadins dans leur expérience de la ville. La naturalité et son impact sur les perceptions seront abordés à travers les variables suivantes : la présence de végétation (dans les espaces souterrains), ses modalités de gestion (dans les espaces verts), l’ambiance sonore (bruits de trafic, bruits de la nature émanant des espaces verts, e.g. le chant d’oiseaux). L’originalité de notre démarche repose sur une double approche :

– Une approche comportementale : elle interroge, à l’aide de questions (ouvertes et fermées), la manière dont l’environnement urbain visuel et sonore impacte la perception de cet environnement (évalué en termes de niveau d’agrément).

– Une approche physiologique : elle s’appuie sur des données physiologiques pour déterminer dans quelle mesure l’environnement urbain visuel et sonore impacte la perception de cet environnement (évalué en termes de modification des réponses physiologiques).

Cofinancé dans le cadre d’une collaboration initiée entre les Labex CeLyA et IMU impliquant les laboratoires EVS, CRNL et LGCB.

Période 2017-2019 - : 24 mois

: Marylise COTTET (EVS)


: Frédéric SEGUR (Grand Lyon, Service Arbres et paysage), Julie VALLET (GRAND LYON, Service Ecologie et développement durable)

: Volodymyr BOGDANOV