会员服务

人脸相关文献代码集锦：人脸检测、人脸识别、人脸生成等

专知出品

     导读    

Github作者ShownX整理了一份人脸相关的资料合集，囊括了人脸检测、对齐、重建、识别、生成等相关的数据集、文献和实现代码等，非常全面，值得收藏。

作者 | ShownX
编译 | Xiaowen
Github：
https://github.com/ShownX/FacePaperCollection


目录

① 工具包

② 人脸检测（Face Detection）

③ 人脸对齐（Face Alignment）

④ 人脸重建（Face Reconstruction）

⑤ 人脸识别（Face Recognition）

⑥ 人脸生成（face Generation）


01

工具包

  • FaRE: Open Source Face Recognition Performance Evaluation Package [Paper]：https://arxiv.org/abs/1901.09447

  • Gluon Toolkit for Face Recognition [MXNET]

  • Deep Learning:

    • MXNet and Gluon: A flexible and efficient library for deep learning.

    • Torch and PyTorch: Tensors and Dynamic neural networks in Python with strong GPU acceleration.

    • TensorFlow: An open-source software library for Machine Intelligence.

    • Caffe and Caffe2: A lightweight, modular, and scalable deep learning framework.

  • Machine Learning:

    • Dlib: A machine learning toolkit.

  • Computer Vision:

    • OpenCV: Open Source Computer Vision Library.

  • Probabilistic Programming

    • Pyro: Deep universal probabilistic programming with Python and PyTorch


02

人脸检测 

Face Detection

数据集

  • Wildest Faces: Face Detection and Recognition in Violent Settings

    • https://arxiv.org/abs/1805.07566

  • WIDER FACE: A Face Detection Benchmark

    • http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/support/paper.pdf

    • Project：http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/WiderFace_Results.html

  • FDDB: Face Detection and Data Set Benchmark

    • https://www.cics.umass.edu/~elm/papers/fddb.pdf

    • Project：

      http://vis-www.cs.umass.edu/fddb/

  • AFLW: Annotated Facial Landmarks in the Wild: A Large-scale, Real-world Database for Facial Landmark Localization

    • http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.384.2988&rep=rep1&type=pdf

    • Project

      https://lrs.icg.tugraz.at/research/aflw/


研究

  • PyramidBox: A Context-assisted Single Shot Face Detector

    • [ Paper] https://arxiv.org/pdf/1803.07737.pdf

    • [TensorFlow] https://github.com/EricZgw/PyramidBox

    • [PyTorch] https://github.com/Goingqs/PyramidBox

    • [MXNet] https://github.com/JJXiangJiaoJun/gluon_PyramidBox

  • Face Attention Network: An Effective Face Detector for the Occluded Faces

    • [Paper] https://arxiv.org/abs/1711.07246

    • [PyTorch] https://github.com/rainofmine/Face_Attention_Network

  • FaceNess-Net: Face Detection through Deep Facial Part Responses:

    • [Paper] https://arxiv.org/pdf/1701.08393.pdf

  • S3FD: Single Shot Scale-invariant Face Detector

    • [Paper] http://openaccess.thecvf.com/content_ICCV_2017/papers/Zhang_S3FD_Single_Shot_ICCV_2017_paper.pdf

    • [Caffe] http://openaccess.thecvf.com/content_ICCV_2017/papers/Zhang_S3FD_Single_Shot_ICCV_2017_paper.pdf

    • [PyTorch] https://github.com/clcarwin/SFD_pytorch

  • Finding Tiny Faces:

    • [Project] https://www.cs.cmu.edu/~peiyunh/tiny/

    • [Paper] https://arxiv.org/abs/1612.04402

    • [MatConvNet + MATLAB] https://github.com/peiyunh/tiny

    • [TensorFlow] https://github.com/cydonia999/Tiny_Faces_in_Tensorflow

    • [MXNET] https://github.com/zzw1123/mxnet-finding-tiny-face

  • SSH: Single Stage Headless Face Detector:

    • [Paper] https://arxiv.org/pdf/1708.03979.pdf

    • [Caffe] https://github.com/mahyarnajibi/SSH

    • [TensorFlow] https://github.com/ailias/Focal-Loss-implement-on-Tensorflow

    • [MXNET] https://github.com/unsky/focal-loss

  • Focal Loss for Dense Object Detection:

    • [Paper] https://arxiv.org/abs/1708.02002

    • [Caffe] https://github.com/chuanqi305/FocalLoss

    • [TensorFlow] https://github.com/ailias/Focal-Loss-implement-on-Tensorflow

    • [MXNET] https://github.com/unsky/focal-loss

  • Face R-CNN:

    • [Paper] https://arxiv.org/abs/1706.01061

    • [Caffe] https://github.com/playerkk/face-py-faster-rcnn

  • FaceBoxes: A CPU Real-time Face Detector with High Accuracy

    • [Paper] http://cn.arxiv.org/abs/1708.05234

    • [Caffe] https://github.com/zeusees/FaceBoxes

  • Multiview Face Detection:

    • [Paper] https://arxiv.org/abs/1502.02766

    • [Caffe] https://github.com/guoyilin/FaceDetection_CNN


03

人脸对齐

Face Alignment

数据集

  • LS3D-W: How far are we from solving the 2D & 3D Face Alignment problem? (and a dataset of 230,000 3D facial landmarks)

    • [Project] https://www.adrianbulat.com/face-alignment

  • AFLW: Annotated Facial Landmarks in the Wild: A Large-scale, Real-world Database for Facial Landmark Localization.

    • [Project] https://lrs.icg.tugraz.at/research/aflw/

  • 300-W

    • [Project] https://ibug.doc.ic.ac.uk/resources/300-W/

  • 300-VW

    • [Project] https://ibug.doc.ic.ac.uk/resources/300-VW/


研究

  • FAN: How far are we from solving the 2D & 3D Face Alignment problem?

    • [Paper] https://arxiv.org/abs/1703.07332

    • [PyTorch] https://github.com/1adrianb/face-alignment

  • JFA: Joint Head Pose Estimation and Face Alignment Framework Using Global and Local CNN Features

    • [Paper] http://cbl.uh.edu/pub_files/07961802.pdf

  • MDM: Mnemonic Descent Method

    • [Paper] https://ibug.doc.ic.ac.uk/media/uploads/documents/trigeorgis2016mnemonic.pdf

    • [TensorFlow] https://github.com/trigeorgis/mdm

  • RDL: Recurrent 3D-2D Dual Learning for Large-pose Facial Landmark Detection

    • [Paper] http://openaccess.thecvf.com/content_ICCV_2017/papers/Xiao_Recurrent_3D-2D_Dual_ICCV_2017_paper.pdf

  • PIFA: Pose-invariant 3D face alignment

    • [Paper] https://arxiv.org/abs/1506.03799

    • [Code] http://cvlab.cse.msu.edu/project-pifa.html


04

人脸重建

Face Reconstruction 


研究

  • UH-E2FAR: End-to-end 3D face reconstruction with deep neural networks:

    • [Paper] https://arxiv.org/abs/1704.05020

  • Multi-View 3D Face Reconstruction with Deep Recurrent Neural Networks:

    • [Paper] http://cbl.uh.edu/pub_files/IJCB-2017-PD.pdf

  • 3D Face Morphable Models "In-the-Wild"

    • [Paper] http://openaccess.thecvf.com/content_cvpr_2017/papers/Booth_3D_Face_Morphable_CVPR_2017_paper.pdf

  • 3DMM-CNN

    • [Paper] https://arxiv.org/pdf/1612.04904.pdf

    • [Code] https://github.com/anhttran/3dmm_cnn

  • VRN

    • [Paper] https://arxiv.org/pdf/1703.07834.pdf

    • [Code] https://github.com/AaronJackson/vrn

    • [Online Demo] http://cvl-demos.cs.nott.ac.uk/vrn

  • 3DFaceNet

    • [Paper] https://arxiv.org/pdf/1708.00980.pdf

  • MoFA: Unsupervised learning for 3D model and pose parameters

    • [Paper] https://arxiv.org/abs/1703.10580

  • 3DMM-STN: Using 3DMM to transfer 2D image to 2D image texture

    • [Paper] https://arxiv.org/abs/1708.07199

  • Dense Semantic and Topological Correspondence of 3D Faces without Landmarks

  • Generating 3D Faces using Convolutional Mesh Autoencoders

    • [Paper] https://arxiv.org/pdf/1807.10267.pdf

    • [Code] https://github.com/anuragranj/coma


05

人脸识别

Face Recognition

教程

  • Deep Learning for Face Recognition

    http://valse.mmcheng.net/deep-learning-for-face-recognition/


数据集

Training sets:

  • MS-Celeb-1M: Microsoft dataset contains around 1M subjects

    • [Project] https://www.microsoft.com/en-us/research/project/ms-celeb-1m-challenge-recognizing-one-million-celebrities-real-world/

    • [Paper] https://arxiv.org/abs/1607.08221

  • CASIA WebFace: 10,575 subjects and 494,414 images

    • [Project] http://www.cbsr.ia.ac.cn/english/CASIA-WebFace-Database.html

    • [Paper] http://arxiv.org/abs/1411.7923

  • CelebA: 202,599 images and 10,177 subjects, 5 landmark locations, 40 binary attributes

    • [Project] http://mmlab.ie.cuhk.edu.hk/projects/

  • VGG-Face2: A large-scale face dataset contains 3.31 million imaes of 9131 identities.

    • [Project] http://www.robots.ox.ac.uk/~vgg/data/vgg_face2/

Face Verification

  • LFW: Labeled Face in the Wild: 13,000 images and 5749 subjects

    • [Download] http://vis-www.cs.umass.edu/lfw/

  • CFP: Celebrities in Frontal-Profile in the Wild

    • [Project] http://www.cfpw.io/

    • [Paper] http://www.cfpw.io/paper.pdf

  • MegaFace: 1 Million Faces for Recognition at Scale, 690,572 subjects

    • [Download] http://megaface.cs.washington.edu/

  • Surveillance Face Recognition Challenge

    • [Project] https://qmul-survface.github.io/

    • [Paper] https://arxiv.org/abs/1804.09691

Face Closed-set Identification

  • UHDB31: UHDB31: A Dataset for Better Understanding Face Recognition across Pose and Illumination Variation

    • [Paper] http://cbl.uh.edu/pub_files/UHDB31_-_CHI_Workshop_-_Final

Face Open-set Identification

  • IJB-C: IARPA Janus Benchmark-C: Face dataset and protocol

    • [Paper] https://noblis.org/wp-content/uploads/2018/03/icb2018.pdf

  • IJB-B: IARPA Janus Benchmark-B Face Dataset

    • [Paper] https://www.nist.gov/document/ijbbchallengedocumentationreadmepdf

  • IJB-A: Pushing the frontiers of unconstrained face detection and recognition: IARPA Janus Benchmark A

    • [Paper] https://www.cv-foundation.org/openaccess/content_cvpr_2015/app/1B_089_ext.pdf

  • Unconstrained Face Detection and Open-Set Face Recognition Challenge

    • [Project] http://vast.uccs.edu/Opensetface/

    • [Paper] https://arxiv.org/abs/1708.02337

  • MegaFace: 1 Million Faces for Recognition at Scale, 690,572 subjects

    • [Download] http://megaface.cs.washington.edu/

Template Generators

Pretrained Models

  • ResNet-101, DenseNet-121 provided by FaRE

    • https://arxiv.org/abs/1901.09447

  • ResNet-50, SE-ResNet-50 provided by VGG-Face2 

    • [Download] https://github.com/ox-vgg/vgg_face2

  • VGG-16 provided by VGG-Face

    • http://www.robots.ox.ac.uk/~vgg/software/vgg_face/

  • InsightFace

    • [Download] https://github.com/deepinsight/insightface

Image-based Template Genearator

  • Pairwise Relation Network, ECCV18:

    • [Paper] https://arxiv.org/pdf/1808.04976.pdf

  • GridFace: Face Rectification via Learning Local Homography Transformation, ECCV18:

    • [Paper] http://openaccess.thecvf.com/content_ECCV_2018/papers/Zhou_GridFace_Face_Rectification_ECCV_2018_paper.pdf

  • Consensus-Driven Propagation in Massive Unlabeled Data for Face Recognition, ECCV18:

    • [Paper] http://openaccess.thecvf.com/content_ECCV_2018/papers/Xiaohang_Zhan_Consensus-Driven_Propagation_in_ECCV_2018_paper.pdf

  • Face Recognition with Contrastive Convolution, ECCV18:

    • [Paper] http://openaccess.thecvf.com/content_ECCV_2018/papers/Chunrui_Han_Face_Recognition_with_ECCV_2018_paper.pdf

  • FaceNet: A Unified Embedding for Face Recognition and Clustering, CVPR15

    • [Paper] https://arxiv.org/abs/1503.03832

    • [TensorFlow] https://github.com/davidsandberg/facenet

  • DeepID series, CVPR14:

    • [DeepID] http://mmlab.ie.cuhk.edu.hk/pdf/YiSun_CVPR14.pdf

    • [DeepID2] http://arxiv.org/abs/1406.4773

    • [DeepID3] http://arxiv.org/abs/1502.00873

  • DeepFace: Closing the Gap to Human-Level Performance in Face Verification, CVPR14:

    • [Paper] https://www.cs.toronto.edu/~ranzato/publications/taigman_cvpr14.pdf

Image-set-based Template Generator

  • Dependency-aware Attention Control for Unconstrained Face Recognition with Image Sets, ECCV, 2018

  • Comparator Network, ECCV, 2018

    • [Pytorch] https://github.com/yeomko22/ComparatorNetwork_pytorch

Training Loss

  • InsightFace (ArcFace): Additive Angular Margin Loss for Deep Face Recognition, ArXiv, 2018

    • [MXNet] https://github.com/deepinsight/insightface

  • CosFace: Large Margin Cosine Loss for Deep Face Recognition, CVPR, 2018

    • [TensorFlow] https://github.com/yule-li/CosFace

    • [MXNet] https://github.com/deepinsight/insightface

  • Ring loss: Convex Feature Normalization for Face Recognition

    • [Paper] https://arxiv.org/abs/1803.00130

    • [PyTorch] https://github.com/Paralysis/ringloss

  • Git Loss for Deep Face Recognition

    • [Paper] https://arxiv.org/abs/1807.08512

  • A-Softmax Loss (SphereFace)

    • [Paper] https://arxiv.org/abs/1704.08063

    • [Caffe] https://github.com/wy1iu/sphereface

  • Triplet Loss

    • [Paper] http://www.cv-foundation.org/openaccess/content_cvpr_2015/app/1A_089.pdf

    • [Torch] https://github.com/cmusatyalab/openface

    • [TensorFlow] https://github.com/davidsandberg/facenet

  • Center Loss

    • [Paper] http://ydwen.github.io/papers/WenECCV16.pdf

    • [Caffe + MATLAB] https://github.com/ydwen/caffe-face

    • [MXNet] https://github.com/pangyupo/mxnet_center_loss

  • Range Loss

    • [Paper] https://arxiv.org/abs/1611.08976

    • [Caffe] https://github.com/Charrin/RangeLoss-Caffe

  • L-Softmax

    • [Paper] https://arxiv.org/abs/1612.02295

    • [Caffe] https://github.com/wy1iu/LargeMargin_Softmax_Loss

    • [MXNet] https://github.com/luoyetx/mx-lsoftmax

  • Marginal Loss

    • [Paper] https://ibug.doc.ic.ac.uk/media/uploads/documents/deng_marginal_loss_for_cvpr_2017_paper.pdf

Face Recognition Pipeline

  • UR2D-E:Evaluation of a 3D-aided Pose Invariant 2D Face Recognition System

    • http://cbl.uh.edu/pub_files/IJCB-2017-XX.pdf

  • SeetaFaceEngine: An open source C++ face recognition engine.

    • [C++] https://github.com/seetaface/SeetaFaceEngine

  • OpenFace: Face recognition with Google's FaceNet deep neural network using Torch]

    • [Torch +Python] https://github.com/cmusatyalab/openface


06

人脸生成

Face Generation

研究

  1. TP-GAN: [Paper] https://arxiv.org/abs/1704.04086

  2. FF-GAN: [Paper] https://arxiv.org/abs/1704.06244

  3. DR-GAN:

    1. [Paper] http://cvlab.cse.msu.edu/pdfs/Tran_Yin_Liu_CVPR2017.pdf

    2. [Website] http://cvlab.cse.msu.edu/project-dr-gan.html

  4. BEGAN: Boundary Equilibrium Generative Adversarial Networks

    1. [Paper] https://arxiv.org/abs/1703.10717


-END-

专 · 知

专知，专业可信的人工智能知识分发，让认知协作更快更好！欢迎登录www.zhuanzhi.ai，注册登录专知，获取更多AI知识资料！

欢迎微信扫一扫加入专知人工智能知识星球群，获取最新AI专业干货知识教程视频资料和与专家交流咨询！

请加专知小助手微信（扫一扫如下二维码添加），加入专知人工智能主题群，咨询技术商务合作~

专知《深度学习:算法到实战》课程全部完成！540+位同学在学习，现在报名，限时优惠！网易云课堂人工智能畅销榜首位！

点击“阅读原文”，了解报名专知《深度学习:算法到实战》课程

展开全文
Top
微信扫码咨询专知VIP会员
Top