An Inclusive Survey of Machine Learning based Hand Gestures Recognition Systems in Recent Applications

Abstract

Abstract Hand gestures represent one of the most prevalent types of body language which can be utilized for interaction and communication. Although the other types of body language represent a more general state of emotional, hand gestures capable of possessing specified linguistic content inside it. Because of the expressiveness and speed in interaction, hand gestures are commonly utilized in human-computer interaction systems (HCI), sign languages, virtual reality, and gaming. In the process of recognizing hand gestures, the complexity and diversity of gestures will extremely impact on the recognition rate and reliability. The existence of machine learning techniques can be effectively exploited in the task of improving the rate of hand gesture recognition. This paper inspected the performance of machine learning techniques in recognizing vision and sensors based hand gestures in the recently existing applications. Additionally, the widely used architecture applied in various datasets has been considered which includes the acquisition of data, pre-processing, the extraction of features, and classification.