基于VGG-Face的人脸识别测试

VGG Face Descriptor 是牛津大学VGG小组的工作，现在已经开源训练好的网络结构和模型参数，本文将基于此模型在caffe上使用自己的人脸数据微调，并进行特征提取与精确度验证。
数据传送门：CASIA WebFace
模型传送门：http://www.robots.ox.ac.uk/~vgg/software/vgg_face/

模型准备

1.从上面的网址中下载VGG-Face已经训练好的模型和网络结构文件，根据deploy.proto文件来修改得到train_val.prototxt文件，主要有：修改数据层的输入和最后全连接层以及损失层等，注意fc8层名称的修改，具体如下：

name: "VGG_FACE_16_layers"
layer {name: "data"type: "Data"top: "data"top: "label"include {phase: TRAIN}transform_param {mirror: truecrop_size: 224
#    mean_file: "data/ilsvrc12/imagenet_mean.binaryproto"}data_param {source: "vggface/webface_train_lmdb"batch_size: 32backend: LMDB}
}
layer {name: "data"type: "Data"top: "data"top: "label"include {phase: TEST}transform_param {mirror: falsecrop_size: 224
#    mean_file: "data/ilsvrc12/imagenet_mean.binaryproto"}data_param {source: "vggface/webface_val_lmdb"batch_size: 32backend: LMDB}
}
......
layer {bottom: "fc7"top: "fc8_s"name: "fc8_s"type: "InnerProduct"param {lr_mult: 10decay_mult: 1}param {lr_mult: 20decay_mult: 0}inner_product_param {num_output: 2031weight_filler {  type: "gaussian"std: 0.01  }  bias_filler {  type: "constant"  value: 0.1}  }
}
layer {name: "accuracy"type: "Accuracy"bottom: "fc8_s"bottom: "label"top: "accuracy"include {phase: TEST}
}
layer {name: "loss"type: "SoftmaxWithLoss"bottom: "fc8_s"bottom: "label"top: "loss"
}

模型训练

本次训练，我采用的是webface中人脸数量在50张以上的个人类别，总共有两千多个。按照caffe的工具转换成lmdb格式即可开始训练。

模型效果测试

模型训练结束之后，在lfw上验证实验效果。lfw数据对使用官方提供的txt文件，不过我觉得格式不太好，就自己用脚本进行了些许修改：
pairs.txt

Abel_Pacheco    1   4
Akhmed_Zakayev  1   3
Akhmed_Zakayev  2   3
Amber_Tamblyn   1   2
Anders_Fogh_Rasmussen   1   3
Anders_Fogh_Rasmussen   1   4

修改后，将一对图片的路径放在一起，属于同一个人为1，否则为0

Abel_Pacheco/Abel_Pacheco_0001.jpg  Abel_Pacheco/Abel_Pacheco_0004.jpg  1
Akhmed_Zakayev/Akhmed_Zakayev_0001.jpg  Akhmed_Zakayev/Akhmed_Zakayev_0003.jpg  1
Akhmed_Zakayev/Akhmed_Zakayev_0002.jpg  Akhmed_Zakayev/Akhmed_Zakayev_0003.jpg  1
Amber_Tamblyn/Amber_Tamblyn_0001.jpg    Amber_Tamblyn/Amber_Tamblyn_0002.jpg    1

脚本如下：

#!/usr/bin/python
#-*- coding: utf-8 -*-
'''
#Created on Thur Mar 2 10:17:38 2017
#Goal:parse pairs.txt of LFW database to label.txt for face #recognition test
#@author: wujiyang
'''
import sysdef get_all_images(filename):file = open(filename)lines = file.readlines()list = []for line in lines:line_split = line.strip('\n').split('\t')if(len(line_split)) == 3:line_split[-1] = line_split[-1].zfill(4)line_split[-2] = line_split[-2].zfill(4)if(len(line_split)) == 4:line_split[-1] = line_split[-1].zfill(4)line_split[-3] = line_split[-3].zfill(4)list.append(line_split)file.close()return listdef save2labelfile(list):file = open('label.txt', 'w')labellines = []for i in range(len(list)):if len(list[i]) == 3:labelline = list[i][0] + '/' + list[i][0] + '_' + list[i][1] + '.jpg' + '\t' + 'original' + '/' + list[i][0] + '/' + list[i][0] + '_' +list[i][2] + '.jpg' + '\t' + '1\n'labellines.append(labelline)elif len(list[i]) == 4:labelline = list[i][0] + '/' + list[i][0] + '_' + list[i][1] + '.jpg' + '\t' + 'original' + '/' + list[i][2] + '/' + list[i][2] + '_' + list[i][3] + '.jpg' + '\t' + '0\n'labellines.append(labelline)file.writelines(labellines)file.close()'''''
使用方法：执行脚本  python pair2label.py pairs.txt
'''
if __name__ == "__main__":if len(sys.argv) != 2:print "Format Error! Usuage: python %s pairs.txt" %(sys.argv[0])sys.exit()list = get_all_images("pairs.txt")save2labelfile(list)print "Done!"

特征提取与精确度验证

此脚本采用caffe的python接口进行特征提取与验证。

# -*- coding: utf-8 -*-
"""
Created on Mon Apr 20 16:55:55 2015@author: wujiyang
@brief：在lfw数据库上验证训练好了的网络
"""
import math
import sklearn
import numpy as np
import matplotlib.pyplot as plt
import skimage
caffe_root = '/home/wujiyang/caffe/'
import sys
sys.path.insert(0, caffe_root + 'python')
import caffeimport sklearn.metrics.pairwise as pw#模型初始化相关操作
def initilize():print 'model initilizing...'deployPrototxt = "./vgg-face-deploy.prototxt"modelFile = "./vgg-face.caffemodel"caffe.set_mode_gpu()caffe.set_device(0)net = caffe.Net(deployPrototxt, modelFile,caffe.TEST)return netdef read_imagelist(labelfile):'''@brief：从列表文件中，读取图像数据到矩阵文件中@param： labelfile 图像列表文件@return ：4D 的矩阵'''file = open(labelfile)lines = file.readlines()test_num=len(lines)file.close()x = np.empty((test_num,3,224,224))y = np.empty((test_num,3,224,224))labels = []i = 0for line in lines:path = line.strip('\n').split('\t')#read left imagefilename = path[0]img = skimage.io.imread(filename,as_grey=False)image = skimage.transform.resize(img,(224,224))*255if image.ndim < 3:print 'gray:'+filenamex[i,0,:,:]=image[:,:]x[i,1,:,:]=image[:,:]x[i,2,:,:]=image[:,:]else:x[i,0,:,:]=image[:,:,0]x[i,1,:,:]=image[:,:,1]x[i,2,:,:]=image[:,:,2]#read right imagefilename = path[1]img = skimage.io.imread(filename,as_grey=False)image = skimage.transform.resize(img,(224,224))*255if image.ndim < 3:print 'gray:'+filenamey[i,0,:,:]=image[:,:]y[i,1,:,:]=image[:,:]y[i,2,:,:]=image[:,:]else:y[i,0,:,:]=image[:,:,0]y[i,1,:,:]=image[:,:,1]y[i,2,:,:]=image[:,:,2]#read labellabels.append(int(path[2]))i=i+1return x, y, labelsdef extractFeature(leftdata,rightdata):#提取左半部分的特征test_num=np.shape(leftdata)[0]#data 是输入层的名字out = net.forward_all(data = leftdata)                                                                    feature1 = np.float64(out['fc7'])featureleft=np.reshape(feature1,(test_num,4096))#np.savetxt('feature1.txt', feature1, delimiter=',')#提取右半部分的特征out = net.forward_all(data = rightdata)                                                                                   feature2 = np.float64(out['fc7'])featureright=np.reshape(feature2,(test_num,4096))#np.savetxt('feature2.txt', feature2, delimiter=',')return featureleft, featurerightdef calculate_accuracy(distance,labels,num):    '''#计算识别率,选取阈值，计算识别率'''    accuracy = {}predict = np.empty((num,))threshold = 0.1while threshold <= 0.9 :for i in range(num):if distance[i] >= threshold:predict[i] = 0else:predict[i] = 1predict_right = 0.0for i in range(num):if predict[i] == labels[i]:predict_right += 1.0current_accuracy = (predict_right / num)accuracy[str(threshold)] = current_accuracythreshold = threshold + 0.001#将字典按照value排序temp = sorted(accuracy.items(), key = lambda d:d[1], reverse = True)highestAccuracy = temp[0][1]thres = temp[0][0]return highestAccuracy, thresif __name__=='__main__':#模型初始化net = initilize()print 'network input :' ,net.inputs  print 'network output： ', net.outputs#读取图像数据leftdata,rightdata,labels = read_imagelist('tmp.txt')#计算特征featureleft, featureright = extractFeature(leftdata, rightdata)#计算每个特征之间的距离 cosine距离test_num = len(labels)mt = pw.pairwise_distances(featureleft, featureright, metric='cosine')distance = np.empty((test_num,))for i in range(test_num):distance[i] = mt[i][i]print 'Distance before normalization:\n', distanceprint 'Distance max:', np.max(distance), ' Distance min:', np.min(distance), '\n'# 距离需要归一化到0--1,与标签0-1匹配distance_norm = np.empty((test_num,))for i in range(test_num):distance_norm[i] = (distance[i]-np.min(distance))/(np.max(distance)-np.min(distance))print 'Distance after normalization:\n', distance_normprint 'Distance_norm max:', np.max(distance_norm), ' Distance_norm min:', np.min(distance_norm), '\n'#根据label和distance_norm计算精确度highestAccuracy, threshold = calculate_accuracy(distance_norm,labels,len(labels))print ("the highest accuracy is : %.4f, and the corresponding threshold is %s \n"%(highestAccuracy, threshold))

基于VGG-Face的人脸识别测试相关推荐

基于深度学习的人脸识别系统（Caffe+OpenCV+Dlib）【三】VGG网络进行特征提取
前言基于深度学习的人脸识别系统,一共用到了5个开源库:OpenCV(计算机视觉库).Caffe(深度学习库).Dlib(机器学习库).libfacedetection(人脸检测库).cudnn(gp ...
基于face_recognition库的摄像头实时人脸识别测试
前言介绍一个基于python的开源人脸识别库,且其离线识别率高达99.38%, github上的网址:github链接该库可以通过python或者命令行即可实现人脸识别的功能.使用dlib深度学习 ...
《繁凡的论文精读》（一）CVPR 2019 基于决策的高效人脸识别黑盒对抗攻击（清华朱军）
点我一文弄懂深度学习所有基础和各大主流研究方向! <繁凡的深度学习笔记>,包含深度学习基础和 TensorFlow2.0,PyTorch 详解,以及 CNN,RNN,GNN,AE,GAN, ...
第十九课.基于sklearn的SVM人脸识别
目录数据集确定人脸的类别标记划分训练集和测试集与训练实验为基于sklearn的SVM人脸识别,使用 SVM 算法对戴眼镜的人脸和不戴眼镜的人脸进行分类,从而完成识别戴眼镜的人脸的任务:实验 ...
基于改进的RPCA人脸识别算法
from:http://www.chinaaet.com/article/3000011311 基于改进的RPCA人脸识别算法作者:首照宇,杨晓帆,莫建文 2015/11/15 18:04:00 摘 ...
基于稀疏表示的人脸识别 (SRC,LASRC,RASL,MRR)
FROM:http://blog.csdn.net/loadstar_kun/article/details/39453839 1. 问题背景信号的稀疏表示并不是新的东西.我们很早就一直在利用这一 ...
基于PYQT编写一个人脸识别软件（2）
前言以前在博客:基于PYQT编写一个人脸识别软件中给出了我自己用PYQT编写的一个小软件.鉴于使用的是开源库--face_recogniton,尽管使用很简单,但是还有些问题,比如:识别黄种人时效 ...
python dlib caffe人脸相似度_基于深度学习的人脸识别系统（Caffe+OpenCV+Dlib）【一】如何配置caffe属性表...
前言基于深度学习的人脸识别系统,一共用到了5个开源库:OpenCV(计算机视觉库).Caffe(深度学习库).Dlib(机器学习库).libfacedetection(人脸检测库).cudnn(gp ...
Python人脸识别教程 - 基于Python的开源人脸识别库：离线识别率高达99.38%
Python人脸识别教程 - 基于Python的开源人脸识别库:离线识别率高达99.38% 仅用 Python 和命令行就可以实现人脸识别的库开源了.该库使用 dlib 顶尖的深度学习人脸识别技术构建 ...
python人脸识别库_基于facenet的实时人脸识别系统
facenet_facerecognition opencv+mtcnn+facenet+python+tensorflow 实现实时人脸识别 Abstract:本文记录了在学习深度学习过程中,使用o ...