The classifier of emotions developed by us allows defining 6 basic emotions: joy, surprise, sadness, anger, fear, disgust. Our internal test showed that, in comparison with foreign analogs, the classifier defines emotions much more precisely.
Recognition of up to 20 informative local features of the face characterizing the psycho-emotional state of a person (ASM)
Calculation of motor units and their classification according to the system of facial movements coding by p. Ekman (FACS Action Units)
Classification of emotions by three independent classifiers: a neural network, a system of decision rules and a weighted sum of features
The ability to determine a neutral facial expression makes the algorithm adaptive to the individual characteristics of each person
Real-time graphics plotting of dynamic changes in the intensity of emotions of the subject and the formation of a report on the results of video processing
The ability to record videos with a webcam, an analysis of all the operating system supported video formats
API for connecting EmoDetect as a module