进入ModelArts
点击如下链接：https://www.huaweicloud.com/product/modelarts.html ， 进入ModelArts主页。点击“立即使用”按钮，输入用户名和密码登录，进入ModelArts使用页面。
创建ModelArts notebook
下面，我们在ModelArts中创建一个notebook开发环境，ModelArts notebook提供网页版的Python开发环境，可以方便的编写、运行代码，并查看运行结果。
第一步：在ModelArts服务主界面依次点击“开发环境”、“创建”
第二步：填写notebook所需的参数：
第三步：配置好notebook参数后，点击下一步，进入notebook信息预览。确认无误后，点击“立即创建”
第四步：创建完成后，返回开发环境主界面，等待Notebook创建完毕后，打开Notebook，进行下一步操作。
在ModelArts中创建开发环境
接下来，我们创建一个实际的开发环境，用于后续的实验步骤。
第一步：点击下图所示的“启动”按钮，加载后“打开”按钮变从灰色变为蓝色后点击“打开”进入刚刚创建的Notebook
第二步：创建一个Python3环境的的Notebook。点击右上角的"New"，然后选择TensorFlow 1.13.1开发环境。
第三步：点击左上方的文件名"Untitled"，并输入一个与本实验相关的名称，如"facial_expression"
在Notebook中编写并执行代码
在Notebook中，我们输入一个简单的打印语句，然后点击上方的运行按钮，可以查看语句执行的结果：
开发环境准备好啦，接下来可以愉快地写代码啦！
数据和代码下载
运行下面代码，进行数据和代码的下载和解压
In [1]:
import os
from modelarts.session import Session
sess = Session()
if sess.region_name == 'cn-north-1':
bucket_path="modelarts-labs/notebook/DL_face_facial_expression/facial_expression.tar.gz"
elif sess.region_name == 'cn-north-4':
bucket_path="modelarts-labs-bj4/notebook/DL_face_facial_expression/facial_expression.tar.gz"
else:
print("请更换地区到北京一或北京四")
if not os.path.exists('./fer2013'):
sess.download_data(bucket_path=bucket_path, path="./facial_expression.tar.gz")
In [2]:
if os.path.exists('./facial_expression.tar.gz'):
# 使用tar命令解压资源包
os.system("tar -xf ./facial_expression.tar.gz")
# 清理压缩包
os.system("rm ./facial_expression.tar.gz")
In [3]:
!pip install mtcnn==0.0.8
!pip install numpy==1.16.2
In [1]:
import numpy as np
import keras
from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D, AveragePooling2D
from keras.layers import Dense, Activation, Dropout, Flatten
from keras.preprocessing.image import ImageDataGenerator
In [2]:
emotions = ('angry', 'disgust', 'fear', 'happy', 'sad', 'surprise', 'neutral')
In [3]:
num_classes = 7 # 类别数
batch_size = 16 # 批大小
epochs = 5 # 训练轮数
In [4]:
with open("./fer2013/fer2013.csv") as f:
content = f.readlines()
lines = np.array(content)
查看一个样本数据
In [5]:
emotion_1, img_1, usage_1 = lines[1].split(",")
In [6]:
val_1 = img_1.split(" ")
pixels_1 = np.array(val_1, 'float32')
print(emotion_1)
print(usage_1)
print(pixels_1.shape)
打印整个数据集的样本数
In [7]:
num_of_instances = lines.size
num_of_instances
Out[7]:
切分训练集和测试集
In [8]:
x_train, y_train, x_test, y_test = [], [], [], []
for i in range(1,num_of_instances):
try:
emotion, img, usage = lines[i].split(",")
val = img.split(" ")
pixels = np.array(val, 'float32')
emotion = keras.utils.to_categorical(emotion, num_classes)
if 'Training' in usage:
y_train.append(emotion)
x_train.append(pixels)
elif 'PublicTest' in usage:
y_test.append(emotion)
x_test.append(pixels)
except:
print("",end="")
数据归一化，并对图像缩放
In [9]:
x_train = np.array(x_train, 'float32')
y_train = np.array(y_train, 'float32')
x_test = np.array(x_test, 'float32')
y_test = np.array(y_test, 'float32')
x_train /= 255
x_test /= 255
x_train = x_train.reshape(x_train.shape[0], 48, 48, 1)
x_train = x_train.astype('float32')
x_test = x_test.reshape(x_test.shape[0], 48, 48, 1)
x_test = x_test.astype('float32')
print(x_train.shape[0], 'train samples')
print(x_test.shape[0], 'test samples')
获取数据生成器
In [10]:
gen = ImageDataGenerator()
train_generator = gen.flow(x_train, y_train, batch_size=batch_size)
In [11]:
def build_model():
model = Sequential()
#1st convolution layer
model.add(Conv2D(64, (5, 5), activation='relu', input_shape=(48,48,1)))
model.add(MaxPooling2D(pool_size=(5,5), strides=(2, 2)))
#2nd convolution layer
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(AveragePooling2D(pool_size=(3,3), strides=(2, 2)))
#3rd convolution layer
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(AveragePooling2D(pool_size=(3,3), strides=(2, 2)))
model.add(Flatten())
#fully connected neural networks
model.add(Dense(1024, activation='relu'))
model.add(Dropout(0.2))
model.add(Dense(1024, activation='relu'))
model.add(Dropout(0.2))
model.add(Dense(num_classes, activation='softmax'))
return model
In [12]:
from keras.models import model_from_json
model = build_model()
model.compile(loss='categorical_crossentropy', optimizer=keras.optimizers.Adam(), metrics=['accuracy'])
model.load_weights('./model/facial_expression_model_weights.h5') # 加载预训练权重
查看模型结构
In [13]:
model.summary()
In [14]:
model.fit_generator(train_generator, steps_per_epoch=batch_size, epochs=epochs)
Out[14]:
In [15]:
import cv2
img = cv2.cvtColor( cv2.imread("./test.jpg"), cv2.COLOR_BGR2RGB)
from PIL import Image
Image.fromarray(img)
Out[15]:
使用mtcnn算法检测出人脸区域
In [16]:
from mtcnn.mtcnn import MTCNN
detector = MTCNN()
result = detector.detect_faces(img)
result
Out[16]:
打印检测出的人脸区域
In [17]:
x,y,w,h = result[0]["box"]
detected_face = img[int(y):int(y+h), int(x):int(x+w)]
detected_face = cv2.cvtColor(detected_face, cv2.COLOR_BGR2GRAY)
detected_face = cv2.resize(detected_face, (48, 48))
Image.fromarray(detected_face)
Out[17]:
使用表情分类模型预测出人脸的表情
In [18]:
from keras.preprocessing import image
import numpy as np
img_pixels = image.img_to_array(detected_face)
img_pixels = np.expand_dims(img_pixels, axis = 0)
img_pixels /= 255
predictions = model.predict(img_pixels)
max_index = np.argmax(predictions[0])
result = emotions[max_index]
result
Out[18]: