A project to teach A.I. understanding human feelings for a better Human Computer interaction. Analyse and simulate Human reactions and responding.
classify emotions inside multimedia
Goal: in the end the A.I. should give feedback about the classified emotion and there reaction inside images, videos, music or even sounds.
for example we’d like to find the most emotional moments in life…
classify and auto cutting of videos,
find stuff that match feelings
google Assistent recognize feelings and interact more personal
add emotion characteristic to the A.I.’s voice and create a ki based avatar to interact more humans like
The first step is to collect training data to capture emotions … so-called „emotion tracking“
We will use the project Tensor-Flow and Google to get the necessary Trainingsdata.
Later we use also YouTube and Webcam to verify our training set.
The Webcam gives the images to recognize the emotions in dependence from the shown videos sequences.
A classification of the Audiovisual data with the corresponding emotions then takes place.
- recognize triggers for certain emotions with this information we hope later to help depressive people by suggesting mood-stabilizing ‚audio-visual content‘.
- helps to recognize attention by emotion tracking
- later help teach people things according to Prof. Dr. Dr. Manfred Spitzer, we reserve ourselves more things like more senses to be addressed, also important for the learning plays the emotion
- filter unknown material from known by their attention
- show persons what there like and make the world more peaceful
- music that makes happy and match the own feelings
recognize spoken words only with the optical recording for disabled persons
#how2
https://www.tensorflow.org/versions/r0.11/get_started/os_setup.html#requirements
https://www.google.de/search?q=happy+man
https://www.google.de/search?q=happy+woman
https://www.google.de/search?q=happy+boy
https://www.google.de/search?q=happy+girl
https://www.google.de/search?q=happy+baby
https://www.google.de/search?q=sad+man
https://www.google.de/search?q=sad+woman
https://www.google.de/search?q=sad+boy
https://www.google.de/search?q=sad+girl
https://www.google.de/search?q=sad+baby
https://www.google.de/search?q=angry+man
https://www.google.de/search?q=angry+woman
https://www.google.de/search?q=angry+boy
https://www.google.de/search?q=angry+girl
https://www.google.de/search?q=angry+baby
