Saved in:
Bibliographic Details
Main Authors: Yang, Jian, Li, Jiakun, Li, Guoming, Shen, Zhen, Wu, Huai-Yu, Fan, Zhaoxin, Huang, Heng
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.16137
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914846802444288
author Yang, Jian
Li, Jiakun
Li, Guoming
Shen, Zhen
Wu, Huai-Yu
Fan, Zhaoxin
Huang, Heng
author_facet Yang, Jian
Li, Jiakun
Li, Guoming
Shen, Zhen
Wu, Huai-Yu
Fan, Zhaoxin
Huang, Heng
contents Multi-view hand mesh reconstruction is a critical task for applications in virtual reality and human-computer interaction, but it remains a formidable challenge. Although existing multi-view hand reconstruction methods achieve remarkable accuracy, they typically come with an intensive computational burden that hinders real-time inference. To this end, we propose MLPHand, a novel method designed for real-time multi-view single hand reconstruction. MLP Hand consists of two primary modules: (1) a lightweight MLP-based Skeleton2Mesh model that efficiently recovers hand meshes from hand skeletons, and (2) a multi-view geometry feature fusion prediction module that enhances the Skeleton2Mesh model with detailed geometric information from multiple views. Experiments on three widely used datasets demonstrate that MLPHand can reduce computational complexity by 90% while achieving comparable reconstruction accuracy to existing state-of-the-art baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2406_16137
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MLPHand: Real Time Multi-View 3D Hand Mesh Reconstruction via MLP Modeling
Yang, Jian
Li, Jiakun
Li, Guoming
Shen, Zhen
Wu, Huai-Yu
Fan, Zhaoxin
Huang, Heng
Computer Vision and Pattern Recognition
Multi-view hand mesh reconstruction is a critical task for applications in virtual reality and human-computer interaction, but it remains a formidable challenge. Although existing multi-view hand reconstruction methods achieve remarkable accuracy, they typically come with an intensive computational burden that hinders real-time inference. To this end, we propose MLPHand, a novel method designed for real-time multi-view single hand reconstruction. MLP Hand consists of two primary modules: (1) a lightweight MLP-based Skeleton2Mesh model that efficiently recovers hand meshes from hand skeletons, and (2) a multi-view geometry feature fusion prediction module that enhances the Skeleton2Mesh model with detailed geometric information from multiple views. Experiments on three widely used datasets demonstrate that MLPHand can reduce computational complexity by 90% while achieving comparable reconstruction accuracy to existing state-of-the-art baselines.
title MLPHand: Real Time Multi-View 3D Hand Mesh Reconstruction via MLP Modeling
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2406.16137