Adapting to the User: A Systematic Review of Personalized Interaction in VR

TL;DR

The paper addresses how VR can be personalized by combining user-state sensing with adaptive interaction. It surveys 132 studies from 2014–2025, identifies a unified five-stage framework (input, modeling, adaptation, system update, feedback), and highlights trends toward multimodal biosignals and AI-driven personalization. Key contributions include a synthesis of technologies, UX effects (presence, embodiment, cybersickness, workload, emotion), and application domains (rehabilitation, mental health, accessibility, education, gaming), as well as implementation challenges (real-time processing, signal noise, privacy) and future directions. The work advances understanding of how to design user-centered, adaptive VR systems and outlines the need for longitudinal evaluations, standardized protocols, and scalable deployments to translate lab results into real-world impact.

Abstract

As virtual reality (VR) systems become increasingly more advanced, they are likewise expected to respond intelligently and adapt to individual user states, abilities, and preferences. Recent work has explored how VR can be adapted and tailored to individual users. However, existing reviews tend to address either user-state sensing or adaptive interaction design in isolation, limiting our understanding of their combined implementation in VR. Therefore, in this paper, we examine the growing research on personalized interaction in VR, with a particular focus on utilizing participants' immersion information and adaptation mechanisms to modify virtual environments and enhance engagement, performance, or a specific goal. We synthesize findings from studies that employ adaptive techniques across diverse application domains and summarize a five-stage conceptual framework that unifies adaptive mechanisms across domains. Our analysis reveals emerging trends, including the integration of multimodal sensors, an increasing reliance on user state inference, and the challenge of balancing responsiveness with transparency. We conclude by proposing future directions for developing more user-centered VR systems.

Figures (7)

• Figure 1: PRISMA flow diagram detailing the study selection process for the systematic review.
• Figure 2: (a) The hardware setup of a multiple-sensorial media platform that provides users with vision, audio, olfaction, and haptic effects according to immersive 360$^{\circ}$ videos sexton2021automatic. Video and audio are played on an Oculus Rift HMD, an Inhalio SBi4v2 olfaction dispenser provides olfaction, and a SteelSeries3 Rival 700 mouse gives haptic feedback. (b) A framework for optimized VR design via immersion level detection raza2024optimized. This system proposes a Polynomial Random Forest for feature generation and is capable of performing classification.
• Figure 3: (a) The percentage distribution of reviewed research papers across different domains. (b) The percentage distribution of reviewed research papers, categorized by the core technical or interactive adaptation technique employed.
• Figure 4: (a) Input modalities (physiological and behavioral data) are fed into the adaptive mechanism powered by AI techniques or rule-based algorithms. Such systems could operate either in real-time or offline. (b) Presence, embodiment, engagement, user perception, and their long-term impacts are several elements that contribute to overall UX. Presence, embodiment, immersion, and engagement are interconnected and interact with one another. User perception refers to how users perceive adaptive elements, whereas long-term impact influences users' performance, retention, and re-engagement. (c) Challenges in current studies include real-time processing ability, signal accuracy and noise, and privacy and security.
• Figure 5: General pipeline for adaptive VR systems for personalized user experiences, including (1) input collection, (2) data processing, (3) adaptive logic, (4) system update, and (5) feedback loop.