Retraction Notice: iMetaTown: A Metaverse System with Multiple Interactive Functions Based on Virtual Reality.

Area of Science:

• Computer Science
• Virtual Reality
• Human-Computer Interaction

Background:

• The Metaverse requires advanced Human-Computer Interaction (HCI) for immersive experiences.
• Realistic 3D face reconstruction is crucial for natural player-player interactions.
• Existing methods often struggle with real-time processing and high-quality 3D alignment.

Purpose of the Study:

• To develop a real-time, immersive Metaverse HCI system using Virtual Reality (VR).
• To enhance player-player interactions through a novel 3D face reconstruction method.
• To optimize 3D model learning and face alignment using 2D images.

Main Methods:

• Implemented a 3D face reconstruction method based on weakly supervised learning.
• Utilized a 2D Self-Supervised Learning (2DASL) approach with 2D face images for 3D model optimization.
• Integrated functional modules for user interactions (hugs, handshakes) and blockchain-based communication (voice, text).
• Developed solutions for managing multiple simultaneous online users.

Main Results:

• The 2DASL face reconstruction method significantly improved 3D face alignment quality.
• The system achieved real-time processing speeds of up to 18 frames per second on a personal computer.
• User feedback indicated high scores for social acceptance, interaction usability, emotional expression, and overall satisfaction.
• The VR HCI system demonstrated robust performance in 3D reconstruction and interaction scenarios.

Conclusions:

• The developed VR HCI system effectively enhances Metaverse interactions through optimized 3D face reconstruction.
• The 2DASL approach provides a viable solution for improving 3D face modeling from 2D images.
• The system meets real-time processing demands and user expectations for immersive virtual environments.