Amélioration de la collecte de l'âge pendant les enquêtes de santé dans les pays à faible revenu / Improving age data collected during health surveys in low-income countries

Description succincte du projet In this project, our goal is to test if an inexpensive and scalable “automatic age estimation” (AAE) method can help improve the measurement of age in LICs. AAE obtains an age estimate from a photograph of an individual by applying facial recognition and machine learning techniques. AAE first constitutes a “training dataset” of face images (i.e., photographs) of individuals with precisely known ages. Then, the “true” age of any individual with unknown (or imprecisely known) age is estimated by statistically matching an image of his/her face to the most closely comparable images in the training dataset. AAE is used in a rapidly growing number of fields including human-machine interaction, online advertising or security systems. But it has not been used to improve demographic estimation in LICs.
Portage administratif
  • Autre
Porteurs du projet (Noms, Institution) Stéphane Helleringer, Université Johns Hopkins
Personne référente du projet au LPED Laurence Fleury
Personnes associées Chong You, René Vidal Laurence Fleury
Pôles associés
  • Tsofi
Bailleurs
  • NIH
Etat du projet Soumis
PDF Modifier