ABSTRACT
Understanding human emotions is a key area of research, since recognizing emotions may provide a plethora of opportunities and applications; for instance, friendlier human–computer interactions with an enhanced communication among humans, by refining emotional intelligence. Hence, human emotion recognition through nonverbal communication can be achieved by capturing body movements. In artificial intelligence research, a machine teaches to detect various patterns using machine learning patterns. Deep learning is a dedicated form of machine learning. The technique that instructs computers to do some operation and behave like humans is done by machine learning. In an unsupervised deep learning approach, the class labels are not required for the learning process. In an unsupervised pretraining learning stage, a backpropagation method is used for fine-tuning. These types of deep learning approaches are used for many applications like object identification, image classification, speech classification, activity, and emotion recognition.
