We conducted an experiment measuring automatically emotions in the face during three n-back tasks with different task difficulties. Often, the different emotions show a fast vacillating, i.e. significant changes of the intensity of the emotion with mostly a low amplitude.
In the literature oscillating or vacillating emotions are in most cases defined as changes between emotions. In papers that report about vacillation of the same emotion often the changes are not described as fast as we have observed (within three seconds or less).
Is there a theory about fast vacillating emotions (i.e. fast changes of its amplitude)?
Do you know good literature with more information (e.g. experiments, theory, observations, ...) about fast vacaillating emotions?
Well, I'm not really into Emotion recognition, but I have been seeing a number of works on something they called "Micro-expressions". It could be that the AUs on the face could cause some transitional emotions which, although not prominent among the conventional six expressions (happy, sad, surprised, afraid etc..), may be highly responsible for this 'seemingly' fast change in the intensity of a particular emotion.
I currently can't lay my hands on those papers talking about Micro-Expressions, but you can just check them up online, I'm sure you'll get good insight. All the best!!!
Hello,
We had done some experiments on objectively quantifying the vacillating emotions that get expressed through facial expressions. You may find this paper to be useful. Let me know. This paper was published in IEEE International Conference on Automatic Face and Gesture Recognition in 2013. I have attached this paper.
Swapna Agararwal
Research Scholar
Indian Statistical Institute
Check https://auduno.github.io/clmtrackr/examples/clm_emotiondetection.html