Labeled Faces in the Wild 人脸识别数据集

Labeled Faces in the Wild Home

New (draft) survey paper:

Labeled Faces in the Wild: A Survey

Erik Learned-Miller, Gary Huang, Aruni RoyChowdhury, Haoxiang Li, Gang Hua

The camera-ready has not yet been submitted. If you see typos or errors, please let us know and we will try to correct them.

New results page:

We have recently updated and changed the format and content of our results page. Please refer to the new
technical report for details of the changes.

Welcome to Labeled Faces in the Wild, a database of face photographs designed for studying the problem of unconstrained face recognition. The data set contains more than 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. The only constraint on these faces is that they were detected by the Viola-Jones face detector. More details can be found in the technical report below.

There are now four different sets of LFW images including the original and three different types of "aligned" images. The aligned images include "funneled images" (ICCV 2007), LFW-a, which uses an unpublished method of alignment, and "deep funneled" images
(NIPS 2012). Among these, LFW-a and the deep funneled images produce superior results for most face verification algorithms over the original images and over the funneled images (ICCV 2007).

Related:

[new] Collected resources related to LFW.

LFW Deep Funneled Images.

LFW attributes file (see Attribute
and Simile Classifiers for Face Verification, Kumar et al.).

Face Detection Data set and Benchmark (FDDB), our new database for face detection research.

Faces in Real-Life Images workshop at the European
Conference on Computer Vision 2008, run by Erik Learned-Miller, Andras Ferencz, and Frederic Jurie.
last updated: 2015/09/22 15:09 EDT

change log

Mailing list:If you wish to receive announcements regarding any changes made to the LFW database, please send email to lfw-subscribe@cs.umass.edu (subject
and body are ignored).
Explore the database:

Alphabetically by first name:

[A)[Alf)[Ang)[B)[Bin)[C)[Che)[Col)[D)[Daw)[Don)[E)[Eri)[F)[G)[Goe)[H) [I)[J)[Jav)[Jes)[Joh)[Jos)[K)[Kim)[L)[Lil)[M)[Mark)[Mel)[Mik)[N)[O)[P) [Per)[Q)[R)[Ric)[Rog)[S)[Sha)[Ste)[T)[Tim)[U)[V)[W)[X)[Y)[Z)
Alphabetically by first name, only people with more than one image:

[A][C][D][E][F][G][H][I][J][K][L][M]
[O][P][Q][R][S][T][U][V][W][X][Y][Z]
Alphabetically by last name:

[A][B][C][D][E][F][G][H][I][J][K][L][M]
[O][P][Q][R][S][T][U][V][W][X][Y][Z]
By number of images per person:

[1 A-E][1
F-J][1 K-O][1
P-T][1 U-Z][2][3][4][5][6-10][11+]
Single page of all names (no thumbnails)

[b]Download the database:

All images as gzipped tar file

(173MB, md5sum a17d05bd522c52d84eca14327a23d494)

[new] All images aligned with deep funneling

(111MB, md5sum 68331da3eb755a505a502b5aacb3c201)

see here for paper to cite.

All images aligned with funneling

(233MB, md5sum 1b42dfed7d15c9b2dd63d5e5840c86ad)

see here for paper to cite.

All images aligned with commercial face alignment software (LFW-a - Taigman, Wolf, Hassner)

See also LFW3D (frontalized LFW images) under LFW resources below.
Superpixel segmentations:

lfw superpixels (328MB, md5sum eb6543ba9bbef54f8ba481c895d3526f)
lfw deep funneled superpixels (129MB, md5sum 5a166aa967e260aa70d55b5785aa7a61)
lfw funneled superpixels (328MB, md5sum f1ede21969d2ad8262a16a26d6212177)

To download LFW attribute values (Attribute and Simile Classifiers for Face Verification, Kumar et al.), see the relevant section
on the results page.
Subset of images - people with name starting with A (14MB)

as zip file
Subset of images - George_W_Bush (individual person with most images) (6.9MB)

as zip file
All names (with number of images for given name) as text file
README - information on file formats and directory structure

Training, Validation, and Testing:
View 1:For development purposes, we recommend using the below training/testing split, which was generated randomly and independently of the splits for 10-fold cross validation, to avoid unfairly overfitting to the sets above during development. For instance, these
sets may be viewed as a model selection set and a validation set. See the tech report below for more details.

Explore the sets: [training][test]

Download the sets: pairsDevTrain.txt, pairsDevTest.txt, peopleDevTrain.txt,peopleDevTest.txt

View 2:As a benchmark for comparison, we suggest reporting performance as 10-fold cross validation using splits we have randomly generated.

Explore the sets: [1][2][3][4][5][6][7][8][9][10]

Download the sets: pairs.txt, people.txt

For information on the file formats, please refer to the README above.

For details on how the sets were created, please refer to the tech report below.

Results:Accuracy and ROC curves for various methods available on results page.
Information:

13233 images
5749 people
1680 people with two or more images

Errata:The following is a list of known errors in LFW. Due to the small number of such errors, the database will be left as is (without corrections) to avoid confusion.

It is important that users of the database provide their algorithms with the databaseas is, i.e. without correcting the errors below, since previous results published for the database did not have the advantage of correcting for these errors.

Currently, there are five incorrectly labeled matched pairs in View 2. While we do not believe this should have a significant effect on accuracy, we do encourage researchers to be aware of these errors when producing any visualizations (e.g. matched pairs most
confidently predicted as mismatched, as the matched pair may actually be mismatched).

The current known errors in View 2 are:

Fold 1: Janica_Kostelic_0001, Janica_Kostelic_0002

Fold 1: Nora_Bendijo_0001, Nora_Bendijo_0002

Fold 5: Jim_OBrien_0001, Jim_OBrien_0002

Fold 5: Jim_OBrien_0001, Jim_OBrien_0003

Fold 5: Elisabeth_Schumacher_0001, Elisabeth_Schumacher_0002

More detail about all the errors is given below.

Note: unless stated otherwise below, any error in a matched pair will mean that the label ("matched") is wrong. Any error in a mismatched pair, even with the person having the wrong identity, will generally be correct (the label of "mismatched" will still be
correct).

Recep_Tayyip_Erdogan_0004 is incorrect (it is an image of Abdullah Gul):



This image appears only in one matched pair in the training set of View 1:

Recep Tayyip Erdogan, 2

Recep Tayyip Erdogan, 4

Anja_Paerson_0001 is incorrect (it is an image of Janica Kostelic):



This image does not appear in a matched or mismatched pair, in either view.

Janica_Kostelic_0001 is incorrect (it is an image of Anja Paerson):



This image appears in one matched pair in the test set of View 1, and the same matched pair and one mismatched pair (with Don_Carcieri_0001) in fold 1 of View 2:

Janica Kostelic, 1

Janica Kostelic, 2

Bart_Hendricks_0001 is incorrect (it is a duplicate image of Ricky_Ray_0001):



This image appears in two mismatched pairs in the training set of View 1, and one mismatched pair in fold 2 of View 2. (None of the mismatched pairs are with Ricky_Ray.)

Carlos_Beltran_0001 is incorrect (it is a duplicate image of Raul_Ibanez_0001):



This image appears in one mismatched pair in the test set of View 1, and one mismatched pair in fold 5 of View 2. (None of the mismatched pairs are with Raul_Ibanez.)

Emmy_Rossum_0001 is incorrect (it is a duplicate image of Eva_Amurri_0001):



This image appears in one mismatched pair in the test set of View 1 (the mismatched pair is not with Eva_Amurri).

Michael_Schumacher_0008 is incorrect (it is an image of Rubens Barrichello):



This image does not appear in a matched or mismatched pair, in either view.

Mahmoud_Abbas_0012 is incorrect (it is an image of Hamad Bin Isa al-Khalifa):



This image does not appear in a matched or mismatched pair, in either view.

Jim_OBrien contains two distinct persons. Specifically, Jim_OBrien_0001 is a different person from Jim_OBrien_0002, Jim_OBrien_0003.

This leads to an error in two matched pairs (0001 with 0002; 0001 with 0003), present in both the training set of View 1 and fold 5 of View 2:

Jim OBrien, 1	Jim OBrien, 2
Jim OBrien, 1	Jim OBrien, 3

John_Gruden_0001 is an incorrect spelling of Jon_Gruden.



This image appears in a mismatched pair in fold 3 of View 2 (not with Jon_Gruden).

Elisabeth_Schumacher contains two distinct persons, where the correct spelling of Elisabeth_Schumacher_0001 is
actually Elizabeth Schumacher. This leads to a incorrect matched pair in both the test set of View 1 and fold 5 of View 2.

Elisabeth Schumacher, 1

Elisabeth Schumacher, 2

Andrew_Gilligan_0001 is actually an image of Andrew_Caldecott.



This image appears in a mismatched pair in the training set of View 1 and in fold 1 of View 2 (neither with Andrew_Caldecott).

Nora_Bendijo_0002 is actually an image of Flor
Montulo, andFlor_Montulo_0002 is actually an image of Nora_Bendijo

Nora Bendijo, 1 correct label	Nora Bendijo, 2 is actually Flor Montulo
Flor Montulo, 1 correct label	Flor Montulo, 2 is actually Nora Bendijo

Nora_Bendijo_0002 appears in an incorrect matched pair in fold 1 of View 2.

Flor_Montulo_0002 appears in an incorrect matched pair in the test set of View 1, and in a mismatched pair in fold 1 of View 2, but not with Nora_Bendijo.

Wang_Yingfan and Yingfan_Wang are
the same person. The two names are never together in a mismatched pair.

Wang_Nan and Nan_Wang are
the same person. The two names are never together in a mismatched pair.

Talisa_Bratt and Talisa_Soto are
the same person.





These images never appear in a mismatched pair together.

Shinya_Taniguchi_0001 is actually an image of Takahiro_Mori.





This image never appears in a mismatched pair with an image of Takahiro_Mori.

Reference:Please cite as:

Gary B. Huang, Manu Ramesh, Tamara
Berg, and Erik Learned-Miller.

Labeled Faces in the Wild: A Database for Studying Face Recognition in Unconstrained Environments.

University of Massachusetts, Amherst, Technical Report 07-49, October, 2007.

[pdf]

BibTeX entry:

@TechReport{LFWTech,
author =       {Gary B. Huang and Manu Ramesh and Tamara Berg and
Erik Learned-Miller},
title =        {Labeled Faces in the Wild: A Database for Studying
Face Recognition in Unconstrained Environments},
institution =  {University of Massachusetts, Amherst},
year =         2007,
number =       {07-49},
month =        {October}}

Gary B. Huang and Erik Learned-Miller.

Labeled Faces in the Wild: Updates and New Reporting Procedures.

University of Massachusetts, Amherst, Technical Report UM-CS-2014-003, May, 2014.

[pdf]

@TechReport{LFWTechUpdate,
author =       {Gary B. Huang Erik Learned-Miller},
title =        {Labeled Faces in the Wild: Updates and New Reporting
Procedures},
institution =  {University of Massachusetts, Amherst},
year =         2014,
number =       {UM-CS-2014-003},
month =        {May}}

LFW funneled images

If you use the LFW images aligned by funneling, please cite:

Gary B. Huang, Vidit Jain, and Erik
Learned-Miller.

Unsupervised joint alignment of complex images.

International Conference on Computer Vision (ICCV), 2007.

@InProceedings{Huang2007a,
author =    {Gary B. Huang and Vidit Jain and Erik Learned-Miller},
title =     {Unsupervised Joint Alignment of Complex Images},
booktitle = {ICCV},
year =      {2007}
}

LFW deep funneled images

If you use the LFW imaged aligned by deep funneling, please cite:

Gary B. Huang, Marwan Mattar, Honglak
Lee, and Erik Learned-Miller.

Learning to Align from Scratch.

Advances in Neural Information Processing Systems (NIPS), 2012.

@InProceedings{Huang2012a,
author =    {Gary B. Huang and Marwan Mattar and Honglak Lee and
Erik Learned-Miller},
title =     {Learning to Align from Scratch},
booktitle = {NIPS},
year =      {2012}
}

Resources:Collected resources related to LFW:

Note: We have not verified the accuracy or reliability of the code and data at the following links; we merely provide them as a convenience. Please use your own judgment about the accuracy of the resources below.

LFWgender

"Getting the known gender based on name of each image in the Labeled Faces in the Wild dataset. This is a python script that calls the genderize.io API with the first name of the person in the image."

CASIA WebFace Database

"While there are many open source implementations of CNN, none of large scale face dataset is publicly available. The current situation in the field of face recognition is that data is more important than algorithm. To solve this problem, we propose a semi-automatical
way to collect face images from Internet and build a large scale dataset containing 10,575 subjects and 494,414 images, called CASIA-WebFace. To the best of our knowledge, the size of this dataset rank second in the literature, only smaller than the private
dataset of Facebook (SCF). We encourage those data-consuming methods training on this dataset and reporting performance on LFW. "

LFW3D - collection of frontalized LFW images and Matlab code for frontalization

"Frontalization is the process of synthesizing frontal facing views of faces appearing in single unconstrained photos. Recent reports have suggested that this process may substantially boost the performance of face recognition systems... we explore the simpler
approach of using a single, unmodified, 3D surface as an approximation to the shape of all input faces. We show that this leads to a straightforward, efficient and easy to implement method for frontalization. More importantly, it produces aesthetic new frontal
views and is surprisingly effective when used for face recognition and gender estimation."

Contact:Questions and comments can be sent to:

Gary Huang - gbhuang@cs.umass.edu

Support:The building of the LFW database was supported by NSF CAREER Award number 0546666.

Change History:
2015/09/22Added Ding et al.*, Ding and Tao*,
and Xu et al.* to results
page.
2015/07/31Updated Errata. Added Baidu, AuthenMetric commercial system results toresults
page.
2015/06/03Updated Errata.
2015/05/19Updated Errata.
2015/04/13Added Sun et al.*, Li and Hua*,
updated betaface.com commercial system*, onresults
page.
2015/01/19Updated Face++ commercial system result*, added betaface.com commercial system result* on results
page.
2014/12/09Added Ouamane et al.* and Sun et al.* to results
page.
2014/12/08Added new LFW-related resources section.

Added Hassner et al.* and TCIT* to results
page.
2014/12/03Added Hu et al.* to results
page.
2014/12/01Added Arashloo and Kittler* to results
page.
2014/09/19Added Li et al.* to results
page.
2014/06/20Added Sun et al.* to results
page.
2014/06/16Added Lu and Tang* to results
page.
2014/06/12Added Sun et al.* to results
page.
2014/05/28Added AUC for two unsupervised methods - LHS and MRF-MLBP - to results page.
2014/05/23Added Zhu et al.* to results
page.
2014/05/22Updated Kumar et al.* with results from journal paper, and added Berg and Belhumeur*,
on results page.
2014/05/21Added Sun et al.* and Hu et al.* to results
page.
2014/05/19Added Taigman et al.* to results
page.
2014/05/09Re-organized results page to more accurately and fairly compare algorithms under a variety of protocols. See the new
technical report for details of the changes.
2014/03/17Added Face++ commercial system result* to results
page.
2014/02/10Added Aurora Computer Services commercial system result* to results
page.
2014/01/09Added VisionLabs commercial system result* to results
page.
2013/12/08Added John_Gruden_0001 labeling error to Errata.
2013/10/18Added Cao et al.*, Cao et al.*,
and Barkan et al.* to results
page.
2013/10/08Added Lei et al.* to results
page.
2013/08/23Added new deep funneled images and corresponding deep funneled superpixels for download. (CJG)
2013/08/12Temporarily disabled deep funneled data set download. (CJG)
2013/08/09Removed superpixel images from database gallery. Added deep funnel aligned images to database gallery. (CJG)
2013/08/08Added deep funneled aligned image set to main page. (CJG)
2013/07/28Added Simonyan et al.* to results
page.
2013/07/13Added Yi et al.* to results
page.
2013/07/13Added Arashloo and Kittler* to results
page.
2013/06/12Added Sharma et al.* to results
page.
2013/05/12Added Zhen et al.* to results
page.
2013/05/02Added Chen et al.* and Chen et al.* to results
page.
2013/04/17Added Anja_Paerson_0001 labeling error to Errata.
2013/03/21Added Li et al.* to results
page.
2012/08/17Added Hussain et al.* to results
page.
2012/07/18Added Berg and Belhumeur* to results
page.
2012/03/03Added Huang et al.* to results
page.
2011/12/14Added Ying and Li* to results
page.
2011/09/07Added link to download computed attribute values for all LFW images produced by Kumar et al., on the results page.
2011/08/28Added Seo and Milanfar* to results
page.
2011/08/08Added images of incorrectly labeled faces, in Errata.
2011/08/08Added Taigman and Wolf* to results
page.
2011/08/01Added Jim_OBrien_0001 labeling error to Errata.
2011/07/18Updated the results page, adding notes on the use of external training data, arranging the image-restricted method results to roughly
reflect the amount of external training data used, and added specific notes on the type of external training data used for each algorithm.
2011/07/12Added Mahmoud_Abbas_0012 labeling error to Errata.
2011/06/28Added Yin et al.* to results
page.
2011/04/28Added superpixel segmentation files to downloads section.
2011/04/04Added Li et al.* to results
page.
2011/01/29Added Pinto and Cox* to results
page.
2010/11/17Added link to related database: Face Detection Data set and Benchmark (FDDB).
2010/10/26Added Nguyen and Bai* to results
page.
2010/09/07Added Michael_Schumacher_0008 labeling error to Errata.
2010/04/17Added Cao et al.* to results
page.
2010/02/08Added Ruiz-del-Solar et al.* and unsupervised (no training data) results toresults
page.
2009/10/26Added Kumar et al.* to results
page.
2009/09/24Added link to LFW-a, LFW images aligned with commercial face alignment software, from Taigman, Wolf, and Hassner, under downloads.
2009/09/02Added Wolf et al.* to results
page.
2009/08/03Added Taigman et al.* to results
page.
2009/07/02Added Guillaumin et al.* to results
page.
2009/06/24Added Carlos_Beltran_0001 and Emmy_Rossum_0001 labeling errors to Errata.
2009/04/02Added Pinto et al.* to results
page.
2009/02/04Added Bart_Hendricks_0001 labeling error to Errata.
2008/07/01Updated LFW technical report with proper reference for VidTIMIT:

C. Sanderson.

Biometric Person Recognition: Face, Speech and Fusion.

VDM-Verlag, 2008.

ISBN 978-3-639-02769-3
2008/06/12Added Errata section and listed two known labeling errors.
2008/02/04Added funneled images and super-pixels images to person pages.

Made all funneled images available as single downloadable file.
2008/01/25Added results page with numbers for method of Nowak and Jurie, CVPR 2007.
2007/11/21Added revised version of technical report.
2007/11/19Added technical report to page.
2007/11/15
Added mailing list and change history to page.

from: http://vis-www.cs.umass.edu/lfw/