EMOTION SCORESΒΆ

Result of emotion recognition based on speech, through a model that considers the following categories:

  • Angry, disgusted, excited, surprised, neutral, sad, anxious, frustrated, frightened, happy

The emotions are ranked by reliability intrinsic to model performance, that is, the angry score is more reliable than the happy score.

The result can be aggregated into groups to simplify the interpretation of the emotional state:

  • negative activating: angry

  • negative deactivating: disgusted, sad, frustrated

  • positive: happy, surprised, anxious, enthusiastic

  • neutral: neutral, frightened

Contains the maximum probability and individual probability of all categories.

The structure contains the following fields:

  • event [string] - indicates the type of object. Valid values:

    • EMOTION RESULT: when the object is the result of emotion recognition.

  • emotion [string] - emotion with greatest probability.

  • p [array] - list of probabilities of each emotional category in emotion [string] format: probability [number]

  • p_groups [array] - list of probabilities of the emotional categories groups in emotion [string] format: probability [number]

Example:

{
"emotion_scores": {
            "event": "EMOTION RESULT",
            "emotion": "frustrado",
            "group": "negativo_desativado",
            "p": {
                "enojado": 0.04615364223718643,
                "frustrado": 0.4350775182247162,
                "triste": 0.27416175603866577,
                "ansioso": 0.036729346960783005,
                "entusiasmado": 0.013698489405214787,
                "feliz": 0.014550963416695595,
                "surpreso": 0.013424476608633995,
                "amedrontado": 0.04797730967402458,
                "neutro": 0.0871482864022255,
                "irritado": 0.03107820823788643
            },
            "p_groups": {
                "negativo_desativado": 0.7553929165005684,
                "positivo": 0.07840327639132738,
                "neutro": 0.13512559607625008,
                "negativo_ativado": 0.03107820823788643
            }
    }
}