Virtual Reality User Experience Classifier Employing Deep Learning For Design Enhancement

Abstract

users in simulated environments. To enhance these experiences, integrating a deep learning-based classifier for evaluating and predicting user behaviour, satisfaction, and engagement is pivotal. This work introduces a Virtual Reality User Experience Classifier utilizing Deep Learning for Design Enhancement. The system offers real-time, data-driven insights into user preferences and behaviours, enabling designers to tailor VR environments effectively. Historically, predicting user experiences in VR was reliant on subjective feedback, manual analysis, or rudimentary computational tools. These methods often failed to capture nuanced user interactions, leading to limited accuracy and inconsistent results. Traditional systems primarily employed surveys, heuristic evaluations, or simple statistical models, which were labor-intensive and lacked scalability. The inability of these approaches to adapt to real-time data or handle large datasets posed significant challenges. The motivation for this research stems from the rapid advancements in AI and VR technologies. The growing demand for personalized, adaptive VR experiences highlighted the need for a robust system that can analyze user interactions holistically. Inspired by the potential of deep learning to uncover hidden patterns in complex data, this research aims to bridge the gap between subjective user feedback and objective system enhancements. The limitations of traditional methods, such as their reliance on static data, human bias, and time-intensive processes, underscore the problem definition. These challenges make it difficult to deliver immersive VR experiences that cater to diverse user needs. Furthermore, the lack of predictive capabilities often results in designs that fail to engage users effectively. The proposed system leverages deep learning algorithms to pre-process data, balance datasets, and classify user interactions accurately. By employing advanced neural networks and models Convolutional Neural Networks (CNNs) and Deep Neural Networks (DNNs), the system predicts user satisfaction with high precision. Additionally, techniques SMOTE for data balancing and loss optimization enhance the model's performance. The outcome is a scalable, efficient, and adaptive system capable of improving VR designs in real time. This novel approach not only enhances user satisfaction but also provides designers with actionable insights, marking a significant step forward in VR design and usability.