awesome-face

🔥 face releated algorithm, datasets and papers 🤔

📝 Paper / Algorithm

2D- Face Recognition

[1] DeepID1 [paper]

Deep Learning Face Representation from Predicting 10,000 Classes

[2] DeepID2 [paper]

Deep Learning Face Representation by Joint Identification-Verification

[3] DeepID2+ [paper]

Deeply learned face representations are sparse, selective, and robust

[4] DeepIDv3 [paper]

DeepID3: Face Recognition with Very Deep Neural Networks

[5] Deep Face [paper]

Deep Face Recognition

[6] Center Loss [paper] [code]

A Discriminative Feature Learning Approach for Deep Face Recognition

[7]Marginal loss [paper]

Marginal loss for deep face recognition

[8] Range Loss[paper]

Range Loss for Deep Face Recognition with Long-tail

[9]Contrastive Loss [paper]

Deep learning face representation by joint identification-verification

[10] FaceNet [paper] [third-party implemention]

FaceNet: A Unified Embedding for Face Recognition and Clustering

[11] NormFace [paper] [code]

NormFace: L2 Hypersphere Embedding for Face Verification

[12] COCO Loss: [paper] [code]

Rethinking Feature Discrimination and Polymerization for Large-scale Recognition

[13] Large-Margin Softmax Loss [paper] [code]

Large-Margin Softmax Loss for Convolutional Neural Networks(L-Softmax loss)

[14]SphereFace： A-Softmax [paper] [code]

SphereFace: Deep Hypersphere Embedding for Face Recognition

[15]AM-Softmax/cosFace [paper AM-Softmax] [paper cosFace] [AM-softmax code]

AM : Additive Margin Softmax for Face Verification

CosFace: Large Margin Cosine Loss for Deep Face Recognition(Tencent AI Lab)

[16] ArcFace: [paper] [code]

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

Face Detection

[1] Cascade CNN [paper] [code]

A Convolutional Neural Network Cascade for Face Detection

[2] MTCNN [Paper] [code]

Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks

[3] ICC - CNN [paper]

Detecting Faces Using Inside Cascaded Contextual CNN

[4] Face R-CNN [Paper]

Face R-CNN

[5] Deep-IR[Paper]

Face Detection using Deep Learning: An Improved Faster RCNN Approach

[6] SSH [paper] [code]

SSH: Single Stage Headless Face Detector

[7] S3FD [paper]

Single Shot Scale-invariant Face Detector

[8] FaceBoxes [paper] [code]

Faceboxes: A CPU Real-time Face Detector with High Accuracy

[9] Scaleface [paper]

Face Detection through Scale-Friendly Deep Convolutional Networks

[10] HR [paper] [code]

Finding Tiny Faces

[11] FAN [paper]

Feature Agglomeration Networks for Single Stage Face Detection.

[12] PyramidBox [paper] [code]

PyramidBox: A Context-assisted Single Shot Face Detector

[13] SRN [paper]

Selective Refinement Network for High Performance Face Detection.

[14] DSFD [paper]

DSFD: Dual Shot Face Detector

[15] VIM FD [paper]

Robust and High Performance Face Detector

[16] ISRN [paper]

Improved Selective Refinement Network for Face Detection

[17] PyramidBox++ [Paper]

PyramidBox++: High Performance Detector for Finding Tiny Face

[18] RetinaFace [paper] [code]

RetinaFace: Single-stage Dense Face Localisation in the Wild

Face Alignment

[1] PRNet [paper] [code]

Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network

[2]LAB Paper [code]

Look at Boundary: A Boundary-Aware Face Alignment Algorithm

[3]PFLD Paper [demo code]

PFLD: A Practical Facial Landmark Detector

[4] 2D & 3D FAN [Paper] [code]

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

Face attack & Defends

[1] A Dataset and Benchmark for Large-Scale Multi-Modal Face Anti-Spoofing

[2] Deep Tree Learning for Zero-Shot Face Anti-Spoofing

[3] Decorrelated Adversarial Learning for Age-Invariant Face Recognition

[4] Multi-Adversarial Discriminative Deep Domain Generalization for Face Presentation Attack Detection

[5] Efficient Decision-Based Black-Box Adversarial Attacks on Face Recognition

⚙ Open source lib

face recognition

face detection

libfaccedetection

📦 Datasets

2D Face Recognition

Datasets	Description	Links	Publish Time
CASIA-WebFace	10,575 subjects and 494,414 images	Download	2014
MegaFace🏅	1 million faces, 690K identities	Download	2016
MS-Celeb-1M🏅	about 10M images for 100K celebrities Concrete measurement to evaluate the performance of recognizing one million celebrities	Download	2016
LFW🏅	13,000 images of faces collected from the web. Each face has been labeled with the name of the person pictured. 1680 of the people pictured have two or more distinct photos in the data set.	Download	2007
VGG Face2🏅	The dataset contains 3.31 million images of 9131 subjects (identities), with an average of 362.6 images for each subject.	Download	2017
UMDFaces Dataset-image	367,888 face annotations for 8,277 subjects.	Download	2016
Trillion Pairs🏅	Train: MS-Celeb-1M-v1c & Asian-Celeb Test: ELFW&DELFW	Download	2018
FaceScrub	It comprises a total of 106,863 face images of male and female 530 celebrities, with about 200 images per person.	Download	2014
Mut1ny🏅	head/face segmentation dataset contains over 17.3k labeled images	Download	2018
IMDB-Face	The dataset contains about 1.7 million faces, 59k identities, which is manually cleaned from 2.0 million raw images.	Download	2018

video face recognition

Datasets	Description	Links	Publish Time
YouTube Face🏅	The data set contains 3,425 videos of 1,595 different people.	Download	2011
UMDFaces Dataset-video🏅	Over 3.7 million annotated video frames from over 22,000 videos of 3100 subjects.	Download	2017
PaSC	The challenge includes 9,376 still images and 2,802 videos of 293 people.	Download	2013
YTC	The data consists of two parts: video clips (1910 sequences of 47 subjects) and initialization data(initial frame face bounding boxes, manually marked).	Download	2008
iQIYI-VID🏅	The iQIYI-VID dataset contains 500,000 videos clips of 5,000 celebrities, adding up to 1000 hours. This dataset supplies multi-modal cues, including face, cloth, voice, gait, and subtitles, for character identification.	Download	2018

3D face recognition

Datasets	Description	Links	Publish Time
Bosphorus🏅	105 subjects and 4666 faces 2D & 3D face data	Download	2008
BD-3DFE	Analyzing Facial Expressions in 3D Space	Download	2006
ND-2006	422 subjects and 9443 faces 3D Face Recognition	Download	2006
FRGC V2.0	466 subjects and 4007 of 3D Face, Visible Face Images	Download	2005
B3D(AC)^2	1000 high quality, dynamic 3D scans of faces, recorded while pronouncing a set of English sentences.	Download	2010

Anti-spoofing

Datasets	# of subj. / # of sess.	Links	Year	Spoof attacks attacks	Publish Time
NUAA	15/3	Download	2010	Print	2010
CASIA-MFSD	50/3	Download(link failed)	2012	Print, Replay	2012
Replay-Attack	50/1	Download	2012	Print, 2 Replay	2012
MSU-MFSD	35/1	Download	2015	Print, 2 Replay	2015
MSU-USSA	1140/1	Download	2016	2 Print, 6 Replay	2016
Oulu-NPU	55/3	Download	2017	2 Print, 6 Replay	2017
Siw	165/4	Download	2018	2 Print, 4 Replay	2018

cross age and cross pose

Datasets	Description	Links	Publish Time
CACD2000	The dataset contains more than 160,000 images of 2,000 celebrities with age ranging from 16 to 62.	Download	2014
FGNet	The dataset contains more than 1002 images of 82 people with age ranging from 0 to 69.	Download	2000
MPRPH	The MORPH database contains 55,000 images of more than 13,000 people within the age ranges of 16 to 77	Download	2016
CPLFW	we construct a Cross-Pose LFW (CPLFW) which deliberately searches and selects 3,000 positive face pairs with pose difference to add pose variation to intra-class variance.	Download	2017
CALFW	Thereby we construct a Cross-Age LFW (CALFW) which deliberately searches and selects 3,000 positive face pairs with age gaps to add aging process intra-class variance.	Download	2017

Face Detection

Datasets	Description	Links	Publish Time
FDDB🏅	5171 faces in a set of 2845 images	Download	2010
Wider-face 🏅	32,203 images and label 393,703 faces with a high degree of variability in scale, pose and occlusion, organized based on 61 event classes	Download	2015
AFW	AFW dataset is built using Flickr images. It has 205 images with 473 labeled faces. For each face, annotations include a rectangular bounding box, 6 landmarks and the pose angles.	Download	2013
MALF	MALF is the first face detection dataset that supports fine-gained evaluation. MALF consists of 5,250 images and 11,931 faces.	Download	2015

Face Attributes

Datasets	Description	Links	Key features	Publish Time
CelebA	10,177 number of identities, 202,599 number of face images, and 5 landmark locations, 40 binary attributes annotations per image.	Download	attribute & landmark	2015
IMDB-WIKI	500k+ face images with age and gender labels	Download	age & gender	2015
Adience	Unfiltered faces for gender and age classification	Download	age & gender	2014
WFLW🏅	WFLW contains 10000 faces (7500 for training and 2500 for testing) with 98 fully manual annotated landmarks.	Download	landmarks	2018
Caltech10k Web Faces	The dataset has 10,524 human faces of various resolutions and in different settings	Download	landmarks	2005
EmotioNet	The EmotioNet database includes950,000 images with annotated AUs. A subset of the images in the EmotioNet database correspond to basic and compound emotions.	Download	AU and Emotion	2017
RAF( Real-world Affective Faces)	29672 number of real-world images, including 7 classes of basic emotions and 12 classes of compound emotions, 5 accurate landmark locations, 37 automatic landmark locations, race, age range and gender attributes annotations per image	Download	Emotions、landmark、race、age and gender	2017

Others

Datasets	Description	Links	Publish Time
IJB C/B/A🏅	IJB C/B/A is currently running three challenges related to face detection, verification, identification, and identity clustering.	Download	2015
MOBIO	bi-modal (audio and video) data taken from 152 people.	Download	2012
BANCA	The BANCA database was captured in four European languages in two modalities (face and voice).	Download	2014
3D Mask Attack	76500 frames of 17 persons using Kinect RGBD with eye positions (Sebastien Marcel).	Download	2013
WebCaricature	6042 caricatures and 5974 photographs from 252 persons collected from the web	Download	2018