dstnation / face-landmarks-detection-benchmark

Face landmarks(fiducial points) detection benchmark

Face-landmarks-detection-benchmark

Face landmarks(fiducial points) detection evaluation.

Attention! It's comparision of specific implementations, not algorithms by itself, so if you know how to improve results let me know.

Name	Rot.	Exp.	Lang	Doc.
Stasm	no	no	C	yes
CLM-framework	?	?	?	?
Dlib	?	?	?	?

Metric:

"The average point-to-point Euclidean error normalized by the inter-ocular distance (measured as the Euclidean distance between the outer corners of the eyes)"
http://ibug.doc.ic.ac.uk/media/uploads/competitions/compute_error.m

"RMSE is very common and is a suitable general-purpose error metric. Compared to the Mean Absolute Error, RMSE punishes large errors"
https://www.kaggle.com/c/facial-keypoints-detection/details/evaluation

To look at:

Kaggle Facial Keypoints Detection
https://github.com/mrgloom/Kaggle-Facial-Keypoints-Detection-Solutions

To try first list:
https://github.com/soundsilence/FaceAlignment

https://github.com/delphifirst/FaceX
https://github.com/ci2cv/face-analysis-sdk  (http://face.ci2cv.net/)
https://github.com/uricamic/flandmark
http://cmp.felk.cvut.cz/~uricamic/flandmark/
http://cmp.felk.cvut.cz/~uricamic/clandmark/
https://github.com/uricamic/clandmark
https://github.com/dnouri/kfkd-tutorial
https://github.com/FaceDetect/jointCascade_py
https://github.com/zhusz/CVPR15-CFSS
http://ibug.doc.ic.ac.uk/resources/fiducial-facial-point-detector-20052007/
http://ibug.doc.ic.ac.uk/resources/facial-point-detector-2010/
https://github.com/kylemcdonald/FaceTracker
http://www.cl.cam.ac.uk/research/rainbow/projects/clmz/
Coarse-to-Fine Auto-Encoder Networks (CFAN) for Real-Time Face Alignment
http://vipl.ict.ac.cn/resources/codes
http://ibug.doc.ic.ac.uk/resources/drmf-matlab-code-cvpr-2013/


ASM/AAM
http://www.milbo.users.sonic.net/stasm/
https://github.com/cxcxcxcx/asmlib-opencv
http://uomasm.sourceforge.net/

constrained local models
https://github.com/TadasBaltrusaitis/CLM-framework

"One Millisecond Face Alignment with an Ensemble of Regression Trees"
http://blog.dlib.net/2014/08/real-time-face-pose-estimation.html
http://www.csc.kth.se/~vahidk/face_ert.html

http://www.ics.uci.edu/~xzhu/face/
https://github.com/TadasBaltrusaitis/CLM-framework


https://github.com/yulequan/face-alignment-in-3000fps
https://github.com/jwyang/face-alignment

http://www.vision.caltech.edu/xpburgos/ICCV13/

Joint Cascade Face Detection and Alignment
https://github.com/luoyetx/JDA

https://github.com/ChrisYang/RCPR
http://www.vision.caltech.edu/xpburgos/ICCV13/

https://github.com/TadasBaltrusaitis/OpenFace

Supervised Descent Method (SDM) for Face Alignment
https://github.com/tntrung/impSDM
https://github.com/patrikhuber/superviseddescent

Not sure 
https://github.com/elador/FeatureDetection
https://github.com/t0nyren/kbdetect

Deep learning:
http://mmlab.ie.cuhk.edu.hk/projects/TCDCN.html
http://mmlab.ie.cuhk.edu.hk/archive/CNN_FacePoint.htm
https://github.com/luoyetx/deep-landmark
https://github.com/zhzhanp/TCDCN-face-alignment
https://github.com/RiweiChen/DeepFace
Recombinator Networks (in theano)
https://github.com/SinaHonari/RCN
https://github.com/ishay2b/VanillaCNN

FANN:
https://github.com/olddocks/facialkeypoints

Javascript:
https://github.com/auduno/clmtrackr

Seems to be commercialized, closed source and not publicly available to download, not worth considering it:
http://www.humansensing.cs.cmu.edu/intraface/

Too simple algorithm, not worth considering it:
https://github.com/sdcoca/facex

Other(blog posts, SO, etc.):

http://www.researchgate.net/post/Which_facial_landmark_detection_tracking_software_is_publically_available_for_research
http://www.learnopencv.com/facial-landmark-detection/

Facial points datasets:

Name	N images	N points	N individuals	Lighting	Age	Race	$	Auth.
MUCT	3755	76	624	yes	yes	yes	no	no

http://www.milbo.org/muct/other-databases.html
[LFPW](http://neerajkumar.org/databases/lfpw/)|1432|29|
[HELEN](http://www.ifp.illinois.edu/~vuongle2/helen/)|2330|192
[AFW]()|?|?
[AFLW](https://lrs.icg.tugraz.at/research/aflw/)|?|?
[IBUG]()|?|68
[PUT]()|?|?
[XM2VTS](http://www.ee.surrey.ac.uk/CVSSP/xm2vtsdb/)|?|?
[ATVS](http://atvs.ii.uam.es/scfacedb_landmarks.html)|?|?|yes
[CACD](http://bcsiriuschen.github.io/CARC/)
[MUG](http://mug.ee.auth.gr/fed/)

Landmark annotation tools:

https://github.com/menpo/landmarker.io

Papers:

"A comparative study of face landmarking techniques"
http://www.busim.ee.boun.edu.tr/~sankur/SankurFolder/Jour_JIVP_Landmarking.pdf
"Supervised Descent Method and its Applications to Face Alignment"
http://www.ri.cmu.edu/pub_files/2013/5/main.pdf
"Deep Convolutional Network Cascade for Facial Point Detection"
http://mmlab.ie.cuhk.edu.hk/archive/CNN/data/CNN_FacePoint.pdf

Other cool benchmarks:

https://github.com/soumith/convnet-benchmarks
https://github.com/DeepMark/deepmark
https://github.com/erikbern/ann-benchmarks
https://github.com/andrewssobral/bgslibrary