BMPCQA: Bioinspired Metaverse Point Cloud Quality Assessment Based on Large Multimodal Models

作者：Huiyu Duan, Kang Fu, Sijing Wu, Yunhao Li, Zicheng Zhang, Qiang Hu, Xiongkuo Min, Guangtao Zhai · 发表于：Advanced Intelligent Systems · 年份：2025 · DOI：10.1002/aisy.202500504 · 被引用次数：3 · 研究领域：Computer Graphics and Visualization Techniques、3D Shape Modeling and Analysis、Remote Sensing and LiDAR Applications

Metaverse has been widely studied in recent years, which aims to build a richer, more interactive, and more integrated digital layer on top of reality. As a fundamental and important 3D representation of metaverse, point clouds have gained significant attention recently and have been applied to many applications. However, assessing the quality of point clouds remains a significant challenge, particularly in scenes involving complex geometries or textures. To this end, this study presents a b ioinspired m etaverse p oint c loud q uality a ssessment (BMPCQA) metric, which simulates the human visual evaluation process to perform the point cloud quality assessment (PCQA) task. Specifically, inspired by local–global perception characteristics of biological visual systems, this study first extracts rendering projection video features, normal image features, and point cloud patch features. Then, these extracted high‐level semantic features and low‐level geometric features are fed into a large multimodal model to achieve biological‐like knowledge feature integration. Extensive experiments on standard PCQA datasets demonstrate that this study's method significantly outperforms existing traditional and learning‐based metrics, correlating more closely with human subjective evaluations. This work paves the way for perceptually aligned and scalable point cloud quality assessment in next‐generation virtual environments. The proposed BMPCQA model is available at: https://github.com/IntMeGro...