As the clinical picture worsens, people with dementia show less and less emotion and are also less and less able to recognize and respond to emotions in others. One of the goals of the Smart Teddy project is to determine whether elderly people with dementia are able to continue to live independently (see more information on the Smart Teddy project page).
One of the indicators for this is the degree to which the patient shows emotions. The aim of this project is to measure this using the sensors that are present in the Smart Teddy and the base station.
The sensors that seem most useful are the microphone, but perhaps the detection of emotions can also be supported by camera images from the base station and motion sensors in the Smart Teddy. Research and data sets exist to recognize emotions from speech. For the time being we assume that if this can be determined on English-language datasets, it can also be done in Dutch speech in a similar way. Existing datasets can be enriched by, for example, putting background sound behind the recordings of authors, adjusting the pitch or cutting certain pieces of audio from the fragments.
- Interest in doing research
- Good English reading and writing skills
- Background in programming or mathematics
If you are interested then please contact one of the leading researchers
Further information: Please visit the page for the Smart Teddy