Emobot is closely linked to the AIMAC (Artificial Intelligence for Multimodal Affective Computing) team at CentraleSupélec, led by Renaud Seguier, co-founder of Emobot.
This team is part of the Image Department of the CNRS 6164 IETR Joint Research Unit.
The AIMAC laboratory already had the ambition to detect depression in the elderly. In 2017, in partnership with Pierre-Paul Vidal (Professor, Director UMR CNRS Cognac-G), they deploy embedded sensors on a hundred elderly people in the city. The goal was to cross-model walking and depression via video processing.
Current work focuses on the detection of stress and micro-expressions in a medical context. The tools developed are mainly based on Artificial Intelligence: GAN, VAE, CNN and auto-encoders.
The main works in progress are:
- The ANR REFLETS project (with Ircam, CentraleSupélec and IRBA: Institut de Recherche Biomédical des Armées) which deals with the therapy on the great depressives of the army (doctors in PTSD). Audio and video processing.
- The correlation between depression and physical activity in children with leukemia undergoing treatment in partnership with CentraleSupélec, the CHU of Rennes and the University of Rennes II. Audio and verbatim processing.
A framework contract links CentraleSupélec and Emobot. The objective is to work on the new version of Emobot.
Our ambition is not only to be able to identify the warning signs of depression but also to propose a solution to avoid depression. We started from the following observation: 3 times more people commit suicide at the age of 75 than at 25. The best way to prevent suicide in the elderly is to fight against isolation (more information here). 4 million people aged 60 and over suffer from the digital divide (more information here).
Communication tools such as sms, emails, WhatsApp, Facebook and other Tiktok promote social link. We propose to bridge the digital divide suffered by isolated people by giving them the possibility to access social networks through speech. Emobot V2 proposes to be the interface between the isolated person and his entourage. Emobot V2 will chat with the person, continue to monitor his emotional state and give him a simple access to learn about the media (emails, sms, video ...) that circulate in his family and feed them.
In our research program with CentraleSupélec, we are evaluating the capacity of Emobot V2 to increase the social link between isolated people and their family, with the aim of getting them out of an incipient depression.