The THERADIA WoZ corpus is an ecological corpus specifically designed for the audiovisual detection of affective states in the healthcare domain. The corpus was built to help the development of a virtual assistant supporting Computerised Cognitive Training (CCT) sessions at home for patients suffering from cognitive impairments. The corpus data come from natural interactions in French involving senior individuals and including both neurotypical and Mild Cognitive Impairment (MCI) participants. The interactions revolve around CCT exercises facilitated by a virtual assistant, operated remotely by a human acting as a Wizard-of-Oz (WoZ). 61 older adults (52 neurotypical seniors and 9 participants with Mild Cognitive Impairment (MCI)) completed cognitive exercises involving memory, language, attention, and planning. The participants’ expressions were fully transcribed and partially annotated based on the dimensions of recent appraisal theories models and 23 affective labels derived from the literature on achievement affects. In total, around 86 hours of interaction data were collected, of which roughly 39.5 hours were segmented and transcribed and 6 hours 20 minutes (2,735 sequences) were selected for detailed affect annotation.

Creator(s)

-

Distributor(s)

ELRA

Right Holder(s)

-