3D人脸重建开源项目整理

最新推荐文章于 2024-04-25 09:48:01 发布

efu莜

最新推荐文章于 2024-04-25 09:48:01 发布

阅读量7.6k

点赞数 19

分类专栏：三维人脸重建论文文章标签： 3D人脸

本文链接：https://blog.csdn.net/qq_35045096/article/details/96498937

版权

三维人脸重建论文专栏收录该内容

2 篇文章 2 订阅

订阅专栏

一、单一视图三维人脸重建

1.End-to-end 3D face reconstruction with deep neural networks
源码：https://github.com/yuefurong97/mxnet-E2FAR

2 Learning Detailed Face Reconstruction from a Single Image
源码：https://github.com/yuefurong97/3DFaceRecon

3.Face Alignment in Full Pose Range: A 3D Total Solution.
源码：https://github.com/yuefurong97/3DDFA

4.Learning-Robust-3D-Face-Reconstruction-and-Discriminative-Identity-Representation（2019.05.16 UESTC）
源码：https://github.com/yuefurong97/SCNN

5.Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision
源码：https://github.com/yuefurong97/RingNet

6.Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set（代码只给出了单一视图人脸重建的部分）
源码：https://github.com/yuefurong97/Deep3DFaceReconstruction

7.Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network (ECCV 2018)
源码：https://github.com/YadiraF/PRNet
https://github.com/yuefurong97/training_codes_for_PRNet_3D_Face
https://github.com/yuefurong97/PRNet_training
https://github.com/liguohao96/pytorch-prnet

8.3D Face Reconstruction from a Single Image using Direct Volumetric CNN Regression
源码：https://github.com/yuefurong97/3D-face-reconstruction

9.MobileFace: 3D Face Reconstruction with Efficient CNN Regression
源码：https://github.com/yuefurong97/MobileFace

10.[Regressing Robust and Discriminative 3D Morphable Models With a Very Deep Neural Network]（2017 CVPR）(https://talhassner.github.io/home/publication/2017_CVPR)
源码：https://github.com/yuefurong97/3dmm_cnn

11.Extreme 3D Face Reconstruction: Seeing Through Occlusions（2018 CVPR）
源码：https://github.com/yuefurong97/extreme_3d_faces

12.SfSNet: Learning Shape, Reflectance and Illuminance of Faces ‘in the wild’（2018 CVPR spotlight）
源码：https://github.com/yuefurong97/SfSNet

13.Large Pose 3D Face Reconstruction from a Single Image via Direct Volumetric CNN Regression （2017 ICCV）
源码：https://github.com/yuefurong97/vrn

14.MoFA: Model-based Deep Convolutional Face Autoencoder for Unsupervised Monocular Reconstruction
源码：https://github.com/waxz/MoFA

15.3D Face Reconstruction with Geometry Details from a Single Image
源码：https://github.com/Juyong/SingleImageReconstruction

16.Disentangled Representation Learning for 3D Face Shape
源码：https://github.com/zihangJiang/DR-Learning-for-3D-Face

17.Unconstrained 3D Face Reconstruction（2015 CVPR）
源码：https://github.com/yuefurong97/CVPR2015-Unconstrained-3D-Face-Reconstruction

18.ExpNet: Landmark-Free, Deep, 3D Facial Expressions
源码：https://github.com/fengju514/Expression-Net

19.Semi-supervised Adversarial Learning to Generate Photorealistic Face Images of New Identities from 3D Morphable Model（2018 ECCV）
源码：https://github.com/barisgecer/facegan

20.Unsupervised Training for 3D Morphable Model Regression（2018 CVPR）
源码：https://github.com/google/tf_mesh_renderer

21.Joint 3D Face Reconstruction and Dense Face Alignment from A Single Image with 2D-Assisted Self-Supervised Learning (2019)
源码：https://github.com/yuefurong97/2DASL

22.Rotate-and-Render: Unsupervised Photorealistic Face Rotation
from Single-View Images（2020 CVPR）
源码：https://github.com/Hangz-nju-cuhk/Rotate-and-Render

23.DF2Net: A Dense-Fine-Finer Network for Detailed 3D Face Reconstruction（2019 ICCV）
源码：https://github.com/xiaoxingzeng/DF2Net

24 Lightweight Photometric Stereo for Facial Details Recovery（2020 CVPR）
源码：https://github.com/Juyong/FacePSNet

25 AvatarMe: Realistically Renderable 3D Facial Reconstruction “in-the-wild”(2020 CVPR)
源码：https://github.com/lattas/AvatarMe（6月份开源代码）

26 Towards High-Fidelity 3D Face Reconstruction from In-the-Wild Images Using
Graph Convolutional Networks (2020 CVPR)
源码：https://github.com/FuxiCV/3D-Face-GCNs

27 Landmark Detection and 3D Face Reconstruction for Caricature using a Nonlinear Parametric Model(2020 CVPR 漫画三维人脸重建)
源码：https://github.com/Juyong/CaricatureFace（有训练代码）

28 Photo-Realistic Facial Details Synthesis from Single Image（2019 ICCV Oral）
附加材料：https://github.com/apchenstu/Facial_Details_Synthesis/blob/master/src/imgs/Supplemental_Material.pdf
源码：https://github.com/apchenstu/Facial_Details_Synthesis

29 Unsup3d (2020 CVPR Oral)
Github:https://github.com/elliottwu/unsup3d

30 GAN2Shape (2021 ICLR Oral)
Github:https://github.com/XingangPan/GAN2Shape

31 Lifting 2D StyleGAN for 3D-Aware Face Generation （2021 CVPR）
Github:暂无

                                           持续更新......

二、多视图三维人脸重建

1.MVF-Net: Multi-View 3D Face Morphable Model Regression
源码：https://github.com/yuefurong97/mvfnet

2.Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set
源码：https://github.com/yuefurong97/Deep3DFaceReconstruction

3D人脸学习：https://github.com/yuefurong97/face3d

四、性能评估器

extreme_3d_faces_evaluator
对应的三维人脸重建项目：
https://github.com/dafuny/extreme_3d_faces

五、OTHERS

1 3ddfa-uv-gan（3DDFA与UV-GAN相结合）
应该是解决3DDFA遮挡情况下的纹理失真的问题
Related:
源码：3DDFA
论文：UV-GAN（2018 CVPR 补全纹理的，目前木有开源？）

2 tf_mesh_renderer （可微分网格渲染器，参看上面的第20篇文章）

3.realtime-2D-to-3D-faces（实时三维人脸重建项目）

4 Hyperparameter-Free Losses for Model-Based Monocular Reconstruction（2019 ICCVW 细节）
源码：https://github.com/hyperparams-free/hyperparams-free-3D-losses（无超参的损失函数进行单眼人脸重建）
Related：
论文：Multi-View 3D Face Reconstruction in the Wild Using Siamese Networks（2019 ICCVW竞赛论文）

5 3D_Face_Reconstrction（一个本科生的个人项目，还是很厉害的~）

6 tinyrenderer（星标8.6k！计算机图形学课程，不使用任何第三方库，展示OPENGL的工作原理，输入：多边形网格+具有纹理的图片输出：渲染模型）

7 DeepFaceRestruction（PRNet的复现代码）

8 Texture-reconstruction（将纹理映射到三维网格的项目好像缺失了一个.h文件）

9 Rasterizer （C++实现具有深度缓冲，纹理映射，法线映射和blinn-phong着色的CPU转发/延迟光栅化器）
<该项目旨在了解渲染算法如何将构成3D世界的一组顶点转换为该世界的2D图像>

六、人脸数据集

1 Now Dataset

efu莜

关注

19
点赞
踩
114

收藏

觉得还不错? 一键收藏
13
评论
3D人脸重建开源项目整理

一、单一视图三维人脸重建1.End-to-end 3D face reconstruction with deep neural networkshttps://github.com/yuefurong97/mxnet-E2FAR2. Learning Detailed Face Reconstruction from a Single Imagehttps://github.com/yu...
复制链接

扫一扫