A SURVEY ON EMOTION DETECTION USING AUDIO DATA PROCESSING AND DEEP LEARNING

The application of speech and audio data is rising day by day. In most of the applications of the speech and audio processing, the Machine Learning (ML) and Artificial Intelligence (AI) play essential role. In ML the audio data analysis is a complex task, due to a number of issues i.e. noise, variations, length, and computational cost. Therefore, audio preprocessing and features extraction is essential. These features are key highlights of the entire data in low level representation. These audio features are further utilized to train a ML algorithm and the trained ML algorithm is used for recognizing the similar audio patterns. In this paper, a review on audio data processing and deep learning based techniques are considered, which are used for identifying the emotion in audio data. Next, the review summary has been prepared and a ML based model proposed to recognize the emotions from audio data. In addition, a brief overview of the presented model has been given. Finally, the expected outcome of the model and future study plan has been discussed.