The emotion challenge dataset contains RGB video set of 50 subjects. For each subject, there are 12 videos representing 6 basic emotions (angry, happy, sad, disgust, contempt, surprise) for real and fake expressions. Each video was recorded with a high resolution camera with 100 frames per second and is about 3-4 seconds. In order for the subjects to express these emotions, they were shown videos which are meant to induce these emotions and were acted accordingly. In each video, subjects started from a neutral emotion and the length of this neutral emotion is not predefined. The sequence of returning to the neutral state was introduced to depict a variation from the acted emotion to the fake emotion to be used in the training process.
The dataset is divided into:
-train set. It contains 480 videos with 40 actors. Each video in train dataset is labeled in the file name (e.g "LABEL.mp4") . Transcriptions to the labels are presented in the following table:
-validation set. It contains 60 videos with 5 actors. Each video name in this set contains an information about the emotion class (Happiness, Sadness, Disgust, Anger, Surprise or Contentment). The labels to this set are provided only after the end of challenge validation phase.
-test set. As a validation set it contains 60 videos with 5 actors. Each video name in this set contains an information about the emotion class (Happiness, Sadness, Disgust, Anger, Surprise or Contentment). The labels for test-set are not provided.