During their first years of life, infants develop physically, emotionally and cognitively at an amazing speed. These early acquisitions lay the foundation for later academic, personal and professional achievements. Many parents are eager to follow this general process and assess how well their child is doing...
We introduce BabyCloud, a platform for capturing, storing and analyzing daylong audio and photo recordings of children's linguistic environments. The platform is composed of 3 components: Baby Logger, a wearable sensor which collects high definition pictures and high-quality recordings to be later decoded by machine learning algorithms, Baby Dock which transfers the data to a secure cloud server and recharges the recorder's batteries and Baby Explorer, a mobile application which allows families to manage and search their data with personalized statistics about their child’s linguistic and cognitive development, games...
Baby Logger is a light, ergonomic and wearable recorder. The device records sounds and conversations through an array of eight high-quality microphones, captures high resolution pictures and data from a three-axis accelerometer. It is protected by an open source hardware licence, and can be reconfigured for particular projects. It is equipped with a large internal memory to allow a maximum of 24 hours of continuous sound recordings and a stream of images every 10 seconds, or short videos.
The operating modes of the device optimize ease of use and respect of privacy. Two versions of the device have been designed, one for the parents and the other for the child. The Baby Logger presents an elegant and child-friendly design, using a snap-on mechanism that allows to easily fix/remove it from any type of clothing or to the infant’s bed.
While the device is recording, no data is being transferred to save on battery life and to protect the child from wave exposure. The wireless data synchronization is performed only when the battery is being recharged onto the Baby Dock. It represents the interface between the wearable recorder and the secure cloud.
The objective of the Baby Explorer is to actively engage families with their dataset by offering an attractive tool with several functionalities: browsing their child's recordings, tracking their child's language and cognitive development with the provided analytics and statistics, protecting and managing the data access...
With the Baby Explorer application, parents can control the audio and image channels (play, rewind, re-scale, etc), filter the data by activity, and collect their infant’s first words! They can bookmark their favorite recordings or pictures and share them on social media.
Dr. Emmanuel Dupoux
Director of the Cognitive Machine Learning team (ENS/CNRS/EHESS/PSL/INRIA) in Paris. He is the creator of BabyCloud and the CTO for the company.
Dr. Xuan-Nga Cao
Linguist with expertise in speech corpora. Has many years of experience in project management. She is the COO for the company.
Malik Ould Arbi
Electronics engineer who graduated from ENSEA and working on prototyping the Baby Logger.
Programmer who graduated from Université Paris-Diderot and working on the application and its functionalities.
Programmer who graduated from Université Paris-Diderot and also working on the application and its functionalities.
Speech therapist providing her expertise on linguistic annotations and speech impairments.
Dr. Paola Guimerans
Creative technologist based in Spain whose academic research is in technology and E-textiles. Joined our team through the WEAR Sustain Network.
Artist and designer, studying in the Art, Media and Technology program at Parsons Paris. Her thesis’ topic is on creating wearable devices.
Want to know more? Contact us!
Find out more about BabyCloud by sending us an email.