Customer service emotion video database construction based on multi-modal scoring and deep learning

The invention provides a customer service emotion video data construction method based on multi-modal scoring and deep learning, and mainly relates to the problem of employee emotion recognition and management in the field of organization behavior research. According to the method, the influence of three kinds of modal information of facial expressions, body languages and voice intonation on emotion recognition is considered, and the accuracy of an emotion recognition result is improved. The method comprises the following steps: firstly, based on an SAM emotion scale, developing an emotion titer amp; a three-level emotion scale of wake-up degree; secondly, starting from three modes of facial expression, limb information and voice intonation, designing an emotion scoring method for two groups of an expert group and a volunteer group; thirdly, writing a deep learning algorithm based on Matlab, fusing emotion scoring results under three modes of facial expression, limb information and voice intonation, and calcu