Skip to content

wangyuan123ac/3DFA-GCN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Learning to detect 3D facial landmarks via heatmap regression with Graph Convolutional Network

by Yuan Wang, Min Cao, Silong Peng and Zhenfeng Fan*

Introduction

This repository is built for the official implementation of:

Learning to detect 3D facial landmarks via heatmap regression with Graph Convolutional Network (AAAI2022)

Overview

Our 3D face alignment model is shown as follows:

3DFA-GCN

We train our model on three publicly available datasets, include BU-3DFE (Yin et al. 2006), FRGCv2.0 (Phillips et al. 2005) and FaceScape (Yang et al. 2020) to demonstrate the effectiveness of the proposed method. Our proposed method achieves 15.1% and 10.6% improvements in terms of the average ME on the BU-3DFE dataset and FRGCv2 dataset respectively. The ME and Std scores by the proposed method reach 1.60 and 1.18 on FaceScape dataset, respectively.

BU-3DFE

FRGCv2

Requirements:

  • Python 3.6.10
  • PyTorch 1.6.0 && torchvision 0.7.0
  • scikit-learn 0.23.2

Acknowledgement

Our code base is partially borrowed from PAConv, DGCNN and PointNet++.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages