Anxiety and Stress Detection through Speech Recognition using CNN

Abstract

Stress is a feeling of emotional tension. It canhaveaninfluenceonourmentalhealth and for the people around us. While anxiety is a natural reaction tostresswhichcanbefearfulthiscanleadtopanicattacks. These mental issues have to be addressed byeveryone.Thispaperexplainshowweareusingvocal/audiodatasettodetectstressandanxietyinaperson.Wehavedevelopedastressandanxietydetection model using deep neural network. Here audio datasets is considered from Kaggle in which the audio consists of 7 emotions i.e., joy, fear, disgust, neutral, sadness, surprised and anger. These audio datasets are used to train and test classification models like CNN. Then the audioispre-processedthroughacousticfeatureextraction, classified through CNN which provides the accuracy based on those 7emotions. By this wecanpredictif the personisstressedorhasanxiety.