1.CNN神经网络
lenet: Gradient-Based Learning Applied to Document Recognition
AlexNet:
2.局部连接的神经网络
2.关于人脸识别方面的论文
facebook cvpr2013: DeepFace: Closing the Gap to Human-Level Performance in Face Verification
3.样本不均衡问题
常见的一些解决样本不均衡的方法包括:
1、 costsensitivelearning
2、 resamplingtechniques
1.“ClassRectification Hard Mining for Imbalanced Deep Learning”
2.“Oversampling forImbalanced Learning Based on K-Means and SMOTE”
3.“IntroducingDeepBalance: Random Deep Belief Network Ensembles to Address Class Imbalance”
3D人脸对齐
References: ------------------- [1] Tal Hassner, Shai Harel, Eran Paz, Roee Enbar, "Effective Face Frontalization in Unconstrained Images," forthcoming. See project page for more details: https://www.openu.ac.il/home/hassner/projects/frontalize [2] T. Hassner, L. Assif, and L. Wolf, "When Standard RANSAC is Not Enough: Cross-Media Visual Matching with Hypothesis Relevancy," Machine Vision and Applications (MVAP), Volume 25, Issue 4, Page 971-983, 2014 Available: https://www.openu.ac.il/home/hassner/projects/poses/ [3] T. Hassner, "Viewing Real-World Faces in 3D," International Conference on Computer Vision (ICCV), Sydney, Austraila, Dec. 2013 Available: https://www.openu.ac.il/home/hassner/projects/poses/ [4] X. Xiong and F. De la Torre, "Supervised Descent Method and its Application to Face Alignment," IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2013 Available: http://www.humansensing.cs.cmu.edu/intraface [5] X. Zhu, D. Ramanan. "Face detection, pose estimation and landmark localization in the wild," Computer Vision and Pattern Recognition (CVPR) Providence, Rhode Island, June 2012. Available: http://www.ics.uci.edu/~xzhu/face/ [6] V. Kazemi, J. Sullivan. "One Millisecond Face Alignment with an Ensemble of Regression Trees," Computer Vision and Pattern Recognition (CVPR), Columbus, Ohio, June, 2014 Available through the dlib library: http://blog.dlib.net/2014/08/real-time-face-pose-estimation.html [7] Yue Wu and Tal Hassner, "Facial Landmark Detection with Tweaked Convolutional Neural Networks," arXiv preprint arXiv:1511.04031, 12 Nov. 2015