嚴重宣告:
利用Python對照片中人臉進行顏值預測!!!
至於結果的可靠性…….
本人概不負責!!!
對結果不滿意或者因為結果分數過低而想不開者,請自行聯絡xxxPh.D.。也就是此顏值預測演演算法的提出者!!!本人只是部分復現了他的演演算法!!!
以上和以下內容純屬玩笑,如有雷同,不甚榮幸。
所需工具
Python版本:3.5.4(64bit)
相關模組:
opencv_python模組、sklearn模組、numpy模組、dlib模組以及一些Python自帶的模組。
環境搭建
(1)安裝相應版本的Python並新增到環境變數中;
(2)pip安裝相關模組中提到的模組。
例如:
若pip安裝報錯,請自行到:
http://www.lfd.uci.edu/~gohlke/pythonlibs/
下載pip安裝報錯模組的whl檔案,並使用:
pip install whl檔案路徑+whl檔案名安裝。
例如:
主要思路
(1)模型訓練
用了PCA演演算法對特徵進行了壓縮降維;
然後用隨機森林訓練模型。
資料源於網路,據說資料“發源地”就是華南理工大學某實驗室,因此我在參考文獻上才加上了這個實驗室的連結。
(2)提取人臉關鍵點
主要使用了dlib庫。
使用官方提供的模型構建特徵提取器。
(3)特徵生成
完全參考了xxxPh.D.的部落格。
(4)顏值預測
利用之前的資料和模型進行顏值預測。
使用方式
有特殊疾病者請慎重嘗試預測自己的顏值,本人不對顏值預測的結果和帶來的所有負面影響負責!!!
言歸正傳。
環境搭建完成後,cmd視窗切換到解壓後的*.py檔案所在目錄。
例如:
開啟test_img檔案夾,將需要預測顏值的照片放入並重新命名為test.jpg。
例如:
若嫌麻煩或者有其他需求,請自行修改:
getLandmarks.py檔案中第13行。
最後依次執行:
train_model.py(想直接用我模型的請忽略此步)
getLandmarks.py
getFeatures.py
Predict.py
train_model.py
# 模型訓練指令碼
import numpy as np
from sklearn import decomposition
from sklearn.ensemble import RandomForestRegressor
from sklearn.externals import joblib
# 特徵和對應的分數路徑
features_path = './data/features_ALL.txt'
ratings_path = './data/ratings.txt'
# 載入資料
# 共500組資料
# 其中前480組資料作為訓練集,後20組資料作為測試集
features = np.loadtxt(features_path, delimiter=',')
features_train = features[0: -20]
features_test = features[-20: ]
ratings = np.loadtxt(ratings_path, delimiter=',')
ratings_train = ratings[0: -20]
ratings_test = ratings[-20: ]
# 訓練模型
# 這裡用PCA演演算法對特徵進行了壓縮和降維。
# 降維之後特徵變成了20維,也就是說特徵一共有500行,每行是一個人的特徵向量,每個特徵向量有20個元素。
# 用隨機森林訓練模型
pca = decomposition.PCA(n_components=20)
pca.fit(features_train)
features_train = pca.transform(features_train)
features_test = pca.transform(features_test)
regr = RandomForestRegressor(n_estimators=50, max_depth=None, min_samples_split=10, random_state=0)
regr = regr.fit(features_train, ratings_train)
joblib.dump(regr, './model/face_rating.pkl', compress=1)
# 訓練完之後提示訓練結束
print('Generate Model Successfully!')
getLandmarks.py
# 人臉關鍵點提取指令碼
import cv2
import dlib
import numpy
# 模型路徑
PREDICTOR_PATH = './model/shape_predictor_68_face_landmarks.dat'
# 使用dlib自帶的frontal_face_detector作為人臉提取器
detector = dlib.get_frontal_face_detector()
# 使用官方提供的模型構建特徵提取器
predictor = dlib.shape_predictor(PREDICTOR_PATH)
face_img = cv2.imread("test_img/test.jpg")
# 使用detector進行人臉檢測,rects為傳回的結果
rects = detector(face_img, 1)
# 如果檢測到人臉
if len(rects) >= 1:
print("{} faces detected".format(len(rects)))
else:
print('No faces detected')
exit()
with open('./results/landmarks.txt', 'w') as f:
f.truncate()
for faces in range(len(rects)):
# 使用predictor進行人臉關鍵點識別
landmarks = numpy.matrix([[p.x, p.y] for p in predictor(face_img, rects[faces]).parts()])
face_img = face_img.copy()
# 使用enumerate函式遍歷序列中的元素以及它們的下標
for idx, point in enumerate(landmarks):
pos = (point[0, 0], point[0, 1])
f.write(str(point[0, 0]))
f.write(',')
f.write(str(point[0, 1]))
f.write(',')
f.write('\n')
f.close()
# 成功後提示
print('Get landmarks successfully')
getFeatures.py
# 特徵生成指令碼
# 具體原理請參見參考論文
import math
import numpy
import itertools
def facialRatio(points):
x1 = points[0]
y1 = points[1]
x2 = points[2]
y2 = points[3]
x3 = points[4]
y3 = points[5]
x4 = points[6]
y4 = points[7]
dist1 = math.sqrt((x1-x2)**2 + (y1-y2)**2)
dist2 = math.sqrt((x3-x4)**2 + (y3-y4)**2)
ratio = dist1/dist2
return ratio
def generateFeatures(pointIndices1, pointIndices2, pointIndices3, pointIndices4, allLandmarkCoordinates):
size = allLandmarkCoordinates.shape
if len(size) > 1:
allFeatures = numpy.zeros((size[0], len(pointIndices1)))
for x in range(0, size[0]):
landmarkCoordinates = allLandmarkCoordinates[x, :]
ratios = []
for i in range(0, len(pointIndices1)):
x1 = landmarkCoordinates[2*(pointIndices1[i]-1)]
y1 = landmarkCoordinates[2*pointIndices1[i] - 1]
x2 = landmarkCoordinates[2*(pointIndices2[i]-1)]
y2 = landmarkCoordinates[2*pointIndices2[i] - 1]
x3 = landmarkCoordinates[2*(pointIndices3[i]-1)]
y3 = landmarkCoordinates[2*pointIndices3[i] - 1]
x4 = landmarkCoordinates[2*(pointIndices4[i]-1)]
y4 = landmarkCoordinates[2*pointIndices4[i] - 1]
points = [x1, y1, x2, y2, x3, y3, x4, y4]
ratios.append(facialRatio(points))
allFeatures[x, :] = numpy.asarray(ratios)
else:
allFeatures = numpy.zeros((1, len(pointIndices1)))
landmarkCoordinates = allLandmarkCoordinates
ratios = []
for i in range(0, len(pointIndices1)):
x1 = landmarkCoordinates[2*(pointIndices1[i]-1)]
y1 = landmarkCoordinates[2*pointIndices1[i] - 1]
x2 = landmarkCoordinates[2*(pointIndices2[i]-1)]
y2 = landmarkCoordinates[2*pointIndices2[i] - 1]
x3 = landmarkCoordinates[2*(pointIndices3[i]-1)]
y3 = landmarkCoordinates[2*pointIndices3[i] - 1]
x4 = landmarkCoordinates[2*(pointIndices4[i]-1)]
y4 = landmarkCoordinates[2*pointIndices4[i] - 1]
points = [x1, y1, x2, y2, x3, y3, x4, y4]
ratios.append(facialRatio(points))
allFeatures[0, :] = numpy.asarray(ratios)
return allFeatures
def generateAllFeatures(allLandmarkCoordinates):
a = [18, 22, 23, 27, 37, 40, 43, 46, 28, 32, 34, 36, 5, 9, 13, 49, 55, 52, 58]
combinations = itertools.combinations(a, 4)
i = 0
pointIndices1 = []
pointIndices2 = []
pointIndices3 = []
pointIndices4 = []
for combination in combinations:
pointIndices1.append(combination[0])
pointIndices2.append(combination[1])
pointIndices3.append(combination[2])
pointIndices4.append(combination[3])
i = i+1
pointIndices1.append(combination[0])
pointIndices2.append(combination[2])
pointIndices3.append(combination[1])
pointIndices4.append(combination[3])
i = i+1
pointIndices1.append(combination[0])
pointIndices2.append(combination[3])
pointIndices3.append(combination[1])
pointIndices4.append(combination[2])
i = i+1
return generateFeatures(pointIndices1, pointIndices2, pointIndices3, pointIndices4, allLandmarkCoordinates)
landmarks = numpy.loadtxt("./results/landmarks.txt", delimiter=',', usecols=range(136))
featuresALL = generateAllFeatures(landmarks)
numpy.savetxt("./results/my_features.txt", featuresALL, delimiter=',', fmt = '%.04f')
print("Generate Feature Successfully!")
Predict.py
# 顏值預測指令碼
from sklearn.externals import joblib
import numpy as np
from sklearn import decomposition
pre_model = joblib.load('./model/face_rating.pkl')
features = np.loadtxt('./data/features_ALL.txt', delimiter=',')
my_features = np.loadtxt('./results/my_features.txt', delimiter=',')
pca = decomposition.PCA(n_components=20)
pca.fit(features)
predictions = []
if len(my_features.shape) > 1:
for i in range(len(my_features)):
feature = my_features[i, :]
feature_transfer = pca.transform(feature.reshape(1, -1))
predictions.append(pre_model.predict(feature_transfer))
print('照片中的人顏值得分依次為(滿分為5分):')
k = 1
for pre in predictions:
print('第%d個人:' % k, end='')
print(str(pre)+'分')
k += 1
else:
feature = my_features
feature_transfer = pca.transform(feature.reshape(1, -1))
predictions.append(pre_model.predict(feature_transfer))
print('照片中的人顏值得分為(滿分為5分):')
k = 1
for pre in predictions:
print(str(pre)+'分')
k += 1
使用演示
作者:佚名
源自:https://www.w3cschool.cn/article/1317851