3D Human Reconstruction from Monocular Vision Based on Neural Fields and Explicit Mesh Optimization
Open Access
- 18 November 2025
- journal article
- research article
- Published by MDPI AG in Electronics
- Vol. 14 (22) , 4512
Abstract
Three-dimensional Human Reconstruction from Monocular Vision is a key technology in Virtual Reality and digital humans. It aims to recover the 3D structure and pose of the human body from 2D images or video. Current methods for dynamic 3D reconstruction of the human body, which are based on monocular views, have low accuracy and remain a challenging problem. This paper proposes a fast reconstruction method based on Instant Human Model (IHM) generation, which achieves highly realistic 3D reconstruction of the human body in arbitrary poses. First, the efficient dynamic human body reconstruction method, InstantAvatar, is utilized to learn the shape and appearance of the human body in different poses. However, due to its direct use of low-resolution voxels as canonical spatial human representations, it is not possible to achieve satisfactory reconstruction results on a wide range of datasets. Next, a voxel occupancy grid is initialized in the A-pose, and a voxel attention mechanism module is constructed to enhance the reconstruction effect. Finally, the Instant Human Model (IHM) method is employed to define continuous fields on the surface, enabling highly realistic dynamic 3D human reconstruction. Experimental results show that, compared to the representative InstantAvatar method, IHM achieves a 0.1% improvement in SSIM and a 2% improvement in PSNR on the PeopleSnapshot benchmark dataset, demonstrating improvements in both reconstruction quality and detail. Specifically, IHM, through voxel attention mechanisms and Mesh adaptive iterative optimization, achieves highly realistic 3D mesh models of human bodies in various poses while ensuring efficiency.Keywords
Funding Information
- National Natural Science Foundation of China (62361041)
- Natural Science Foundation of Jiangxi Province (20242BAB25053)
This publication has 46 references indexed in Scilit:
- Surfel-Based Gaussian Inverse Rendering for Fast and Relightable Dynamic Human Reconstruction From Monocular VideosIEEE Transactions on Pattern Analysis and Machine Intelligence, 2025
- StruGauAvatar: Learning Structured 3D Gaussians for Animatable Avatars From Monocular VideosIEEE Transactions on Visualization and Computer Graphics, 2025
- SAT-NGP : Unleashing Neural Graphics Primitives for Fast Relightable Transient-Free 3D Reconstruction From Satellite Imagery2024
- Deep Learning for 3D Reconstruction, Augmentation, and Registration: A Review PaperEntropy, 2024
- Fast-SNARF: A Fast Deformer for Articulated Neural FieldsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
- Consistent 3D Hand Reconstruction in Video via Self-Supervised LearningIEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
- NeRFCommunications of the ACM, 2021
- Modular primitives for high-performance differentiable renderingACM Transactions on Graphics, 2020
- Lightweight Photometric Stereo for Facial Details RecoveryarXiv, 2020
- DeepHuman: 3D Human Reconstruction from a Single ImagearXiv, 2019