Voice and Speech Emotion Recogntion
The Emotion API takes a voice, speech or conversation in an audio or video as an input, and return the confidence across a set of emotions for the audio.
The emotions are given in two ways: discrete and dimensional. The discrete emotions detected are happiness, neutral, sadness, anger, fear, disgust, contempt, and surprise. These emotions are understood to be cross-culturally. The dimensional emotion are detected by arousal and valance. This metric are widely accepted and in emotion research community.
User Scenario: Chatbot, Smart living, Robot, Advertising, Entertainment, Customer Service.