Actions
Forum Thread Details
  • Replies 3 replies
  • Subscribers 345 subscribers
  • Views 1912 views
  • Users 0 members are here
Related

Types of sensors to detect human emotions?

nuriag

Hi,

 

This might sound a little bit weird, but I'm a Masters Architecture student who is currently designing a populist city in which monuments would have some sort of sensors that would read human emotions around them, to find out whether people approved or not of these monuments.

 

I was told to write a technological thesis on sensors that could be placed in this monuments and read emotions, needles to say, I don't know much about sensors (yet). I imagine there are six major things that we can try to read to detect emotions:

 

  • Facial expression
  • Respiratory rate
  • People can be detected remotely by their charge or capacitance.
  • Heart-rate
  • Galvanic skin response
  • Blood pressure

 

Imagining that a monument would be placed in a public outdoor area, what would be the solution you think would suit this best? I imagine I will have to use different types of sensors all at once, including cameras, etc. but I would really appreciate your (far more educated on the topic) solutions!

 

Thanks in advance, and I apologise if the question is too weird!

  • Hi Nuria,

     

    I don't have an answer for you, but it's not a weird question.. such determination of human state would be useful in other industries too. Perhaps your research will need to do surveys so you can compare your captured and processed data to see how accurate it is. I was wondering, maybe for the application you have in mind, maybe the emotion per individual isn't as necessary as an average. Maybe an image recognition or otherwise tracking system (e.g. signals from a mobile phone, or an app and BLE beacon inside the monument) will provide you with a sample of information from thousands of visitors, so you can see how long they linger near a monument, as another datapoint.

    On the face of it, it is difficult because this is outdoors, so it becomes harder to know if a person's state is due to other factors (e.g. cold weather, running to see the monument due to lack of time, or an argument with someone). Some AI will likely be needed to pull out patterns.  Temperature isn't on your list, but that can be sensed too (e.g. thermal array) and might have useful data for your needs.

    In summary, although it looks difficult, it is still worth a try - you never know, it could be possible to extract information that was previously difficult. I've been caught out on occasion by incorrect assumptions that things are too difficult/inaccurate before, so nowadays I'm not so sceptical : ) Sometimes algorithms/processing and modern sensors can turn conventional ideas up-side down.

    Also, maybe your research could also explore work in other situations like airports, where it is necessary to know if a person is calm or excited, or if they have a fever due to illness. That work may have findings useful for your requirement.

  • A few other parameters you might find useful are:

    • pupil dilation
    • posture
    • gaze direction and duration (I expect these would be most useful for your study)
    • gestures
    • audio