Detection of Emotions from Speech using Deep Learning Techniques and Traditional Techniques: A Survey

Detection of emotions from speech data is an important but exigent task of Human-Computer Communication. Speech Emotion Recognition system is a hybrid model that analyses and detect emotions embedded in speech signals. Emotion detection is a prominent task as emotions are instinctive in nature. Recently, various technologies are used to extract emotions from speech databases including traditional approaches for Speech Emotion Recognition (SER). As a substitute to traditional techniques in SER, Deep Learning techniques is great weapon to deal with the challenges in SER. This paper presents a brief review on various Deep Learning techniques and throws light on some latest literature relating to speech-based emotion recognition system.