Hand Gesture Detection and Classification Accuracy Improvement in NN and Clustering Segmentation Hybrid Methods

Abstract

When it comes to human-computer interaction, gesture recognition is crucial. In hand gesture recognition, modified convolutional neural network (CNN) models are used for hand posture prediction, hand motion capture, and hand object interaction. We used these models to estimate hand poses on a data set, motion capture accessible postures dataset, and hand object interaction on a dataset. Hand motion capture accuracy is improved by analyzing such components.