2019.01.06
2018比较新的人脸关键点
https://blog.csdn.net/daniaokuye/article/details/82748599#commentBox
FAN:时间5s
DAN:时间1.5s
openpose face:openpose主打的是2d人体姿态的估计,face只是其中一部分,不作为重点,点少,效果差,速度慢
Vanilla CNN:
VANface除了侧脸效果不好之外,其他都相当的好,精度和速度的完美结合,输入图片是60*60,网络也不是很深。
https://talhassner.github.io/home/publication/2017_TPAMI_2 (作者的官网)
https://github.com/ishay2b/VanillaCNN (网络原型代码)
下面四个链接内容,模型或者用的都是第一个,或者都是来源于Vanilla CNN
https://github.com/lsy17096535/face-landmark (轻量化网络,i7-7700 GPU-1060, 6ms+)
https://github.com/cooparation/VanillaCNN_faceLandmark (face-landmark同一个模型)
https://github.com/cunjian/face_alignment(i7-7700 GPU-1060, 2ms+,有做网络优化)
look at boundary: https://github.com/wywu/LAB
face-landmark:https://github.com/jiangwqcooler/face-landmark-localization (暂时不清楚用的什么网络,,i7-7700 GPU-1060, 6ms+)
2018.10.18
级联MobileNet-V2实现CelebA人脸关键点检测(附训练源码)
https://www.ctolib.com/topics-130493.html
深度学习人脸关键点检测方法----综述
https://blog.csdn.net/u011995719/article/details/78890333
--------------------------------------------------------------------
!以下内容为2015年及2015年以前的陈旧内容 !
--------------------------------------------------------------------
转自:https://www.douban.com/note/525032729/
https://github.com/delphifirst/FaceX
https://github.com/ci2cv/face-analysis-sdk
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://mmlab.ie.cuhk.edu.hk/archive/CNN_FacePoint.htm
http://mmlab.ie.cuhk.edu.hk/projects/TCDCN.html
http://ibug.doc.ic.ac.uk/resources/fiducial-facial-point-detector-20052007/
http://ibug.doc.ic.ac.uk/resources/facial-point-detector-2010/
http://www.humansensing.cs.cmu.edu/intraface/
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
check if it detects points or only align face
https://github.com/jwyang/face-alignment
https://github.com/soundsilence/FaceAlignment
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
Deep learning:
https://github.com/olddocks/caffe-facialkp
http://corpocrat.com/2015/02/24/facial-keypoints-extraction-using-deep-learning-with-caffe/
http://danielnouri.org/notes/2014/12/17/using-convolutional-neural-nets-to-detect-facial-keypoints-tutorial/
FANN:
https://github.com/olddocks/facialkeypoints
Kaggle:
https://www.kaggle.com/c/facial-keypoints-detection/data
Javascript:
https://github.com/auduno/clmtrackr
Too simple algorithm, not worth considering it:
https://github.com/sdcoca/facex