人臉識別[一] 演算法和資料庫總結

Face-Resources

Following is a growing list of some of the materials I found on the web for research on face recognition algorithm。

Papers

1。 ［DeepFace］（

https：//www。

cs。toronto。edu/~ranzato

/publications/taigman_cvpr14。pdf

）。A work from Facebook。

2。 ［FaceNet］（

http：//www。

cv-foundation。org/opena

ccess/content_cvpr_2015/app/1A_089。pdf

）。A work from Google。

3。 ［ One Millisecond Face Alignment with an Ensemble of Regression Trees］（

http：//www。

csc。kth。se/~vahidk/pape

rs/KazemiCVPR14。pdf

）。 Dlib implements the algorithm。

4。 ［DeepID］（

http：//

mmlab。ie。cuhk。edu。hk/pd

f/YiSun_CVPR14。pdf

）

5。 ［DeepID2］（［1406。4773］ Deep Learning Face Representation by Joint Identification-Verification）

6。 ［DeepID3］（Face Recognition with Very Deep Neural Networks）

7。 ［Learning Face Representation from Scratch］（［1411。7923］ Learning Face Representation from Scratch）

8。 ［Face Search at Scale： 80 Million Gallery］（80 Million Gallery）

9。 ［A Discriminative Feature Learning Approach for Deep Face Recognition］（

http：//

ydwen。github。io/papers/

WenECCV16。pdf

）

10。 ［NormFace： L2 Hypersphere Embedding for Face Verification］（

https：//

arxiv。org/abs/1704。0636

9

）。* attention： model released ！*

11。 ［SphereFace： Deep Hypersphere Embedding for Face Recognition］（Deep Hypersphere Embedding for Face Recognition）

12。［VGGFace2： A dataset for recognising faces across pose and age ］A dataset for recognising faces across pose and age

Datasets

1。 ［CASIA WebFace Database］（Center for Biometrics and Security Research）。 10，575 subjects and 494，414 images

2。 ［Labeled Faces in the Wild］（

http：//

vis-www。cs。umass。edu/lf

w/

）。13，000 images and 5749 subjects

3。 ［Large-scale CelebFaces Attributes （CelebA） Dataset］（403 Forbidden） 202，599 images and 10，177 subjects。 5 landmark locations， 40 binary attributes。

4。 ［MSRA-CFW］（MSRA-CFW： Data Set of Celebrity Faces on the Web - Microsoft Research）。 202，792 images and 1，583 subjects。

5。 ［MegaFace Dataset］（MegaFace） 1 Million Faces for Recognition at Scale

690，572 unique people

6。 ［FaceScrub］（vintage - resources）。 A Dataset With Over 100，000 Face Images of 530 People。

7。 ［FDDB］（FDDB ： Main）。Face Detection and Data Set Benchmark。 5k images。

8。 ［AFLW］（ICG - Research）。Annotated Facial Landmarks in the Wild： A Large-scale， Real-world Database for Facial Landmark Localization。 25k images。

9。 ［AFW］（Face Detection Matlab Code）。 Annotated Faces in the Wild。 ~1k images。

10。［3D Mask Attack Dataset］（3D Mask Attack Dataset）。 76500 frames of 17 persons using Kinect RGBD with eye positions （Sebastien Marcel）

11。 ［Audio-visual database for face and speaker recognition］（MOBIO - DDP）。Mobile Biometry MOBIO

http：//www。

mobioproject。org/

12。 ［BANCA face and voice database］（The BANCA Database）。 Univ of Surrey

13。 ［Binghampton Univ 3D static and dynamic facial expression database］（

http：//www。

cs。binghamton。edu/~liju

n/Research/3DFE/3DFE_Analysis。html

）。 （Lijun Yin， Peter Gerhardstein and teammates）

14。 ［The BioID Face Database］（BioID Face Database | Dataset for Face Detection | facedb - BioID）。 BioID group

15。 ［Biwi 3D Audiovisual Corpus of Affective Communication］（ETHZ - Computer Vision Lab：）。 1000 high quality， dynamic 3D scans of faces， recorded while pronouncing a set of English sentences。

16。 ［Cohn-Kanade AU-Coded Expression Database］（The Affect Analysis Group at Pittsburgh）。 500+ expression sequences of 100+ subjects， coded by activated Action Units （Affect Analysis Group， Univ。 of Pittsburgh。

17。 ［CMU/MIT Frontal Faces ］（CBCL SOFTWARE）。 Training set： 2，429 faces， 4，548 non-faces； Test set： 472 faces， 23，573 non-faces。

18。 ［AT&T Database of Faces］（The Database of Faces） 400 faces of 40 people （10 images per people）

Trained Model

1。 ［openface］（cmusatyalab/openface）。 Face recognition with Google‘s FaceNet deep neural network using Torch。

2。 ［VGG-Face］（VGG Face Descriptor）。 VGG-Face CNN descriptor。 Impressed embedding loss。

3。 ［SeetaFace Engine］（seetaface/SeetaFaceEngine）。 SeetaFace Engine is an open source C++ face recognition engine， which can run on CPU with no third-party dependence。

4。 ［Caffe-face］（ydwen/caffe-face） - Caffe Face is developed for face recognition using deep neural networks。

5。 ［Norm-Face］（happynear/NormFace） - Norm Face， finetuned from ［center-face］（ydwen/caffe-face） and ［Light-CNN］（AlfredXiangWu/face_verification_experiment）

6。 ［VGG-Face2］VGG-Face 2Dataset

Software

1。 ［OpenCV］（OpenCV library）。 With some trained face detector models。

2。 ［dlib］（dlib C++ Library - Machine Learning）。 Dlib implements a state-of-the-art of face Alignment algorithm。

3。 ［ccv］（liuliu/ccv）。 With a state-of-the-art frontal face detector

4。 ［libfacedetection］（ShiqiYu/libfacedetection）。 A binary library for face detection in images。

5。 ［SeetaFaceEngine］（seetaface/SeetaFaceEngine）。 An open source C++ face recognition engine。

Frameworks

1。 ［Caffe］（Caffe | Deep Learning Framework）

2。 ［Torch7］（torch/torch7）

3。 ［Theano］（Welcome - Theano 1。0。0 documentation）

4。 ［cuda-convnet］（

https：//

code。google。com/p/cuda-

convnet/

）

5。 ［MXNET］（apache/incubator-mxnet）

6。 ［Tensorflow］（tensorflow）

7。 ［tiny-dnn］（tiny-dnn/tiny-dnn）

Miscellaneous

1。 ［faceswap］（matthewearl/faceswap） Face swapping with Python， dlib， and OpenCV

2。 ［Facial Keypoints Detection］（Facial Keypoints Detection | Kaggle） Competition on Kaggle。

3。 ［An implementation of Face Alignment at 3000fps via Local Binary Features］（freesouls/face-alignment-at-3000fps）

layout： post

category： deep_learning

title： Face Recognition

date: 2015-10-09

Papers

DeepID

Deep Learning Face Representation from Predicting 10,000 Classes

intro： CVPR 2014

paper：

http：//

mmlab。ie。cuhk。edu。hk/pd

f/YiSun_CVPR14。pdf

github：

https：//

github。com/stdcoutzyx/D

eepID_FaceClassify

DeepID2

Deep Learning Face Representation by Joint Identification-Verification

paper：

http：//

papers。nips。cc/paper/54

16-analog-memories-in-a-balanced-rate-based-network-of-e-i-neurons

基於Caffe的DeepID2實現

1。

http：//www。

miaoerduo。com/deep-lear

ning/%E5%9F%BA%E4%BA%8Ecaffe%E7%9A%84deepid2%E5%AE%9E%E7%8E%B0%EF%BC%88%E4%B8%8A%EF%BC%89。html

2。

http：//www。

miaoerduo。com/deep-lear

ning/%E5%9F%BA%E4%BA%8Ecaffe%E7%9A%84deepid2%E5%AE%9E%E7%8E%B0%EF%BC%88%E4%B8%AD%EF%BC%89。html

3。

http：//www。

miaoerduo。com/deep-lear

ning/%E5%9F%BA%E4%BA%8Ecaffe%E7%9A%84deepid2%E5%AE%9E%E7%8E%B0%EF%BC%88%E4%B8%8B%EF%BC%89。html

DeepID2+

Deeply learned face representations are sparse, selective, and robust

arxiv：

http：//

arxiv。org/abs/1412。1265

video：

http：//

research。microsoft。com/

apps/video/？id=260023

mirror：

http：//

pan。baidu。com/s/1boufl3

x

MobileID

MobileID: Face Model Compression by Distilling Knowledge from Neurons

intro： AAAI 2016 Oral。 CUHK

intro： MobileID is an extremely fast face recognition system by distilling knowledge from DeepID2

project page：

http：//

personal。ie。cuhk。edu。hk

/~lz013/projects/MobileID。html

paper：

http：//

personal。ie。cuhk。edu。hk

/~pluo/pdf/aaai16-face-model-compression。pdf

github：

https：//

github。com/liuziwei7/mo

bile-id

DeepFace

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

intro： CVPR 2014。 Facebook AI Research

paper：

https：//www。

cs。toronto。edu/~ranzato

/publications/taigman_cvpr14。pdf

slides：

http：//

valse。mmcheng。net/ftp/2

0141126/MingYang。pdf

github：

https：//

github。com/RiweiChen/De

epFace

Deep Face Recognition

intro： BMVC 2015

paper：

http：//www。

robots。ox。ac。uk/~vgg/pu

blications/2015/Parkhi15/parkhi15。pdf

homepage：

http：//www。

robots。ox。ac。uk/~vgg/so

ftware/vgg_face/

github（Keras）：

https：//

github。com/rcmalli/kera

s-vggface

FaceNet

FaceNet: A Unified Embedding for Face Recognition and Clustering

intro： Google Inc。 CVPR 2015

arxiv：

http：//

arxiv。org/abs/1503。0383

2

github（Tensorflow）：

https：//

github。com/davidsandber

g/facenet

github（Caffe）：

https：//

github。com/hizhangp/tri

plet

Real time face detection and recognition

intro： Real time face detection and recognition base on opencv/tensorflow/mtcnn/facenet

github：

https：//

github。com/shanren7/rea

l_time_face_recognition

Targeting Ultimate Accuracy: Face Recognition via Deep Embedding

intro： CVPR 2015

arxiv：

http：//

arxiv。org/abs/1506。0731

0

Learning Robust Deep Face Representation

arxiv：

https：//

arxiv。org/abs/1507。0484

4

A Light CNN for Deep Face Representation with Noisy Labels

arxiv：

https：//

arxiv。org/abs/1511。0268

3

github：

https：//

github。com/AlfredXiangW

u/face_verification_experiment

Pose-Aware Face Recognition in the Wild

paper： www。cv-foundation。org/openaccess/content_cvpr_2016/papers/Masi_Pose-Aware_Face_Recognition_CVPR_2016_paper。pdf

Triplet Probabilistic Embedding for Face Verification and Clustering

intro： Oral Paper in BTAS 2016； NVIDIA Best paper Award

arxiv：

https：//

arxiv。org/abs/1604。0541

7

github（Keras）：

https：//

github。com/meownoid/fac

e-identification-tpe

Recurrent Regression for Face Recognition

arxiv：

http：//

arxiv。org/abs/1607。0699

9

A Discriminative Feature Learning Approach for Deep Face Recognition

intro： ECCV 2016

intro： center loss

paper：

http：//

ydwen。github。io/papers/

WenECCV16。pdf

github：

https：//

github。com/ydwen/caffe-

face

github：

https：//

github。com/pangyupo/mxn

et_center_loss

Deep Face Recognition with Center Invariant Loss

intro： ACM MM Workshop

paper：

http：//

www1。ece。neu。edu/~yuewu

/files/2017/twu024。pdf

How Image Degradations Affect Deep CNN-based Face Recognition?

arxiv：

http：//

arxiv。org/abs/1608。0524

6

VIPLFaceNet: An Open Source Deep Face Recognition SDK

keywords： VIPLFaceNet / SeetaFace Engine

arxiv：

http：//

arxiv。org/abs/1609。0389

2

SeetaFace Engine

intro： SeetaFace Engine is an open source C++ face recognition engine， which can run on CPU with no third-party dependence。

github：

https：//

github。com/seetaface/Se

etaFaceEngine

A Discriminative Feature Learning Approach for Deep Face Recognition

intro： ECCV 2016

paper：

http：//

ydwen。github。io/papers/

WenECCV16。pdf

Sparsifying Neural Network Connections for Face Recognition

paper：

http：//www。

ee。cuhk。edu。hk/~xgwang/

papers/sunWTcvpr16。pdf

Range Loss for Deep Face Recognition with Long-tail

arxiv：

https：//

arxiv。org/abs/1611。0897

6

Hybrid Deep Learning for Face Verification

intro： TPAMI 2016。 CNN+RBM

paper：

http：//www。

ee。cuhk。edu。hk/~xgwang/

papers/sunWTpami16。pdf

Towards End-to-End Face Recognition through Alignment Learning

intro： Tsinghua University

arxiv：

https：//

arxiv。org/abs/1701。0717

4

Multi-Task Convolutional Neural Network for Face Recognition

arxiv：

https：//

arxiv。org/abs/1702。0471

0

NormFace: L2 Hypersphere Embedding for Face Verification

arxiv：

https：//

arxiv。org/abs/1704。0636

9

github：

https：//

github。com/happynear/No

rmFace

SphereFace: Deep Hypersphere Embedding for Face Recognition

intro： CVPR 2017

arxiv：

http：//

wyliu。com/papers/LiuCVP

R17。pdf

github：

https：//

github。com/wy1iu/sphere

face

demo：

http：//

v-wb。youku。com/v_show/i

d_XMjk3NTc1NjMxMg==。html

L2-constrained Softmax Loss for Discriminative Face Verification

https：//

arxiv。org/abs/1703。0950

7

Low Resolution Face Recognition Using a Two-Branch Deep Convolutional Neural Network Architecture

intro： Amirkabir University of Technology & MIT

arxiv：

https：//

arxiv。org/abs/1706。0624

7

Enhancing Convolutional Neural Networks for Face Recognition with Occlusion Maps and Batch Triplet Loss

https：//

arxiv。org/abs/1707。0792

3

Model Distillation with Knowledge Transfer in Face Classification, Alignment and Verification

https：//

arxiv。org/abs/1709。0292

9

Improving Heterogeneous Face Recognition with Conditional Adversarial Networks

https：//

arxiv。org/abs/1709。0284

8

Face Sketch Matching via Coupled Deep Transform Learning

intro： ICCV 2017

arxiv：

https：//

arxiv。org/abs/1710。0291

4

Additive Margin Softmax for Face Verification

keywords： additive margin Softmax （AM-Softmax），

arxiv：

https：//

arxiv。org/abs/1801。0559

9

github：

https：//

github。com/happynear/AM

Softmax

Face Recognition via Centralized Coordinate Learning

https：//

arxiv。org/abs/1801。0567

8

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

arxiv：

https：//

arxiv。org/abs/1801。0769

8

github：

https：//

github。com/deepinsight/

insightface

CosFace: Large Margin Cosine Loss for Deep Face Recognition

https：//

arxiv。org/abs/1801。0941

4

Ring loss: Convex Feature Normalization for Face Recognition

intro： CVPR 2018

arxiv：

https：//

arxiv。org/abs/1803。0013

0

Pose-Robust Face Recognition via Deep Residual Equivariant Mapping

intro： CVPR 2018。 CUHK & SenseTime Research

arxiv：

https：//

arxiv。org/abs/1803。0083

9

Video Face Recognition

Attention-Set based Metric Learning for Video Face Recognition

https：//

arxiv。org/abs/1704。0380

5

SeqFace: Make full use of sequence information for face recognitio

arxiv：

https：//

arxiv。org/abs/1803。0652

4

github：

https：//

github。com/huangyangyu/

SeqFace

Facial Point / Landmark Detection

Deep Convolutional Network Cascade for Facial Point Detection

homepage：

http：//

mmlab。ie。cuhk。edu。hk/ar

chive/CNN_FacePoint。htm

paper：

http：//www。

ee。cuhk。edu。hk/~xgwang/

papers/sunWTcvpr13。pdf

github：

https：//

github。com/luoyetx/deep

-landmark

Facial Landmark Detection by Deep Multi-task Learning

intro： ECCV 2014

project page：

http：//

mmlab。ie。cuhk。edu。hk/pr

ojects/TCDCN。html

paper：

http：//

personal。ie。cuhk。edu。hk

/~ccloy/files/eccv_2014_deepfacealign。pdf

github（Matlab）：

https：//

github。com/zhzhanp/TCDC

N-face-alignment

A Recurrent Encoder-Decoder Network for Sequential Face Alignment

intro： ECCV 2016 oral

project page：

https：//

sites。google。com/site/x

ipengcshomepage/eccv2016

arxiv：

https：//

arxiv。org/abs/1608。0547

7

slides：

https：//

drive。google。com/file/d

/0B-FLp_bljv_1OTVrMF9OM21IbW8/view

github：

https：//

github。com/xipeng13/rec

urrent-face-alignment

RED-Net: A Recurrent Encoder-Decoder Network for Video-based Face Alignment

intro： IJCV

arxiv：

https：//

arxiv。org/abs/1801。0606

6

Detecting facial landmarks in the video based on a hybrid framework

arxiv：

http：//

arxiv。org/abs/1609。0644

1

Deep Constrained Local Models for Facial Landmark Detection

arxiv：

https：//

arxiv。org/abs/1611。0865

7

Effective face landmark localization via single deep network

arxiv：

https：//

arxiv。org/abs/1702。0271

9

A Convolution Tree with Deconvolution Branches: Exploiting Geometric Relationships for Single Shot Keypoint Detection

https：//

arxiv。org/abs/1704。0188

0

Deep Alignment Network: A convolutional neural network for robust face alignment

intro： CVPRW 2017

arxiv：

https：//

arxiv。org/abs/1706。0178

9

gihtub：

https：//

github。com/MarekKowalsk

i/DeepAlignmentNetwork

Joint Multi-view Face Alignment in the Wild

https：//

arxiv。org/abs/1708。0602

3

FacePoseNet: Making a Case for Landmark-Free Face Alignment

https：//

arxiv。org/abs/1708。0751

7

Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networks

https：//

arxiv。org/abs/1711。0675

3

Brute-Force Facial Landmark Analysis With A 140,000-Way Classifier

intro： AAAI 2018

arxiv：

https：//

arxiv。org/abs/1802。0177

7

github：

https：//

github。com/mtli/BFFL

Style Aggregated Network for Facial Landmark Detection

intro： CVPR 2018

arxiv：

https：//

arxiv。org/abs/1803。0410

8

github：

https：//

github。com/D-X-Y/SAN

Deep Adaptive Attention for Joint Facial Action Unit Detection and Face Alignment

https：//

arxiv。org/abs/1803。0558

8

Projects

Using MXNet for Face-related Algorithm

github：

https：//

github。com/tornadomeet/

mxnet-face

clmtrackr: Javascript library for precise tracking of facial features via Constrained Local Models

github：

https：//

github。com/auduno/clmtr

ackr

blog：

http：//

auduno。com/post/6188827

7175/fitting-faces

demo：

http：//

auduno。github。io/clmtra

ckr/examples/facesubstitution。html

demo：

http：//

auduno。github。io/clmtra

ckr/face_deformation_video。html

demo：

http：//

auduno。github。io/clmtra

ckr/examples/clm_emotiondetection。html

demo：

http：//

auduno。com/post/8421458

7523/twisting-faces

DeepLogo

intro： A brand logo recognition system using deep convolutional neural networks。

github：

https：//

github。com/satojkovic/D

eepLogo

Deep-Leafsnap

intro： LeafSnap replicated using deep neural networks to test accuracy compared to traditional computer vision methods。

github：

https：//

github。com/sujithv28/De

ep-Leafsnap

FaceVerification: An Experimental Implementation of Face Verification, 96.8% on LFW

github：

https：//

github。com/happynear/Fa

ceVerification

InsightFace

intro： Face Recognition Project on MXnet

arxiv：

https：//

github。com//deepinsight

/insightface

OpenFace

OpenFace: Face Recognition with Deep Neural Networks

homepage：

http：//

cmusatyalab。github。io/o

penface/

github：

https：//

github。com/cmusatyalab/

openface

github：

https：//

github。com/aybassiouny/

OpenFaceCpp

OpenFace 0.2.0: Higher accuracy and halved execution time

homepage：

http：//

bamos。github。io/2016/01

/19/openface-0。2。0/

OpenFace: A general-purpose face recognition library with mobile applications

paper：

http：//

reports-archive。adm。cs。cmu。edu

/anon/anon/usr0/ftp/2016/CMU-CS-16-118。pdf

OpenFace: an open source facial behavior analysis toolkit

intro： a state-of-the art open source tool intended for facial landmark detection， head pose estimation，

facial action unit recognition， and eye-gaze estimation。

github：

https：//

github。com/TadasBaltrus

aitis/OpenFace

Resources

Face-Resources

github：

https：//

github。com/betars/Face-

Resources

Created by ruyiwei on 10/04/2018。