Adaptive Deep Learning-Based Interactive Virtual Digital Human System for Enhanced Customer Engagement in the Metaverse

This study presents a unique framework called the Adaptive Deep Learning-based Interactive Virtual Digital Human System (ADL-IVDHS), which combines intelligent algorithms and system modeling to enhance customer engagement in sales domains. With the help of virtual digital humans who can participate in individualized interactions, the metaverse provides businesses with innovative chances to improve consumer engagement. To transform sales processes into complete virtual worlds, this study presents an ADL-IVDHS. The system combines intelligent interaction, behavior prediction, and real-time multi-object recognition to produce interactive digital representations that can reflect human speech and behavior. Some of the advanced features in our proposed approach are fusing the inputs of RGB and LiDAR to obtain an accurate vision of the digital environments. Additionally, we combine the advantages of the natural language processing (NLP) technique to improve customer interaction. This helps to improve the customer engagement and the digital interaction rates. Also, we implement the advantages of YOLOv5 and improved Kalman filters for tracking and detecting objects to guarantee effective interactions in changing digital environments. A simulation of the suggested study is performed under MATLAB and TensorFlow to prove the effectiveness of the proposed model in terms of real-time customer engagement achievement, response time, and reduction of false positives; this shows an improvement in customer satisfaction and sales outcomes.