Gesture recognition is still developing rapidly and is a major part of human-computer interactions (HCI), which aims to provide users with more intuitive and natural ways of communication with the computer. The traditional methods of HCI, such as keyboards, mice, and touchscreens, have their own set of limitations, especially, when it comes to the more flexible hardware. Gesture recognition in general and hand gesture recognition in particular, has been a viable modality of non-contact human-computer interaction (HCI) with the devices. This research focuses on machine learning-based methods for the control of hand gestures in the execution of common tasks gestures in the execution of common tasks of hand gestures as the execution of the most common tasks, i.e., changing pages, scrolling, and doing other interactive functions, is performed. Here, machine learning algorithms are utilized along with sensor data to train the system to identify user hand movements in real-time. Through this method, users are enabled to interact with the application in a fast and easy way. Their technical problems along with possible improvements for performance and user experience of the gesture-based HCI systems are also addressed in the paper.
Ibrahim et al. (Thu,) studied this question.