第三部分 模型：卷积神经网络
这一部分让我们从代码开始：
import keras
import pandasas pd
import numpyas np
from kerasimport Sequential
from keras.layersimport Conv2D, Activation, MaxPooling2D, Dropout, Flatten, Dense
from keras.modelsimport load_model
from matplotlibimport pyplot
TRAIN_FILE ='training.csv'
TEST_FILE ='test.csv'
SAVE_PATH ='model'
VALIDATION_SIZE =100 #验证集大小
EPOCHS =100 #迭代次数
BATCH_SIZE =64 #每个batch大小，稍微大一点的batch会更稳定
EARLY_STOP_PATIENCE =10 #控制early stopping的参数
def input_data(test=False):
file_name = TEST_FILEif testelse TRAIN_FILE
df = pd.read_csv(file_name)
cols = df.columns[:-1]
#dropna()是丢弃有缺失数据的样本，这样最后7000多个样本只剩2140个可用的。
df = df.dropna()
df['Image'] = df['Image'].apply(lambda img: np.fromstring(img, sep=' ') /255.0)
X = np.vstack(df['Image'])
X = X.reshape((-1,96,96,1))
if test:
y =None
else:
y = df[cols].values /96.0 #将y值缩放到[0,1]区间
return X, y
#最后生成提交结果的时候要用到
keypoint_index = {
'left_eye_center_x':0,
'left_eye_center_y':1,
'right_eye_center_x':2,
'right_eye_center_y':3,
'left_eye_inner_corner_x':4,
'left_eye_inner_corner_y':5,
'left_eye_outer_corner_x':6,
'left_eye_outer_corner_y':7,
'right_eye_inner_corner_x':8,
'right_eye_inner_corner_y':9,
'right_eye_outer_corner_x':10,
'right_eye_outer_corner_y':11,
'left_eyebrow_inner_end_x':12,
'left_eyebrow_inner_end_y':13,
'left_eyebrow_outer_end_x':14,
'left_eyebrow_outer_end_y':15,
'right_eyebrow_inner_end_x':16,
'right_eyebrow_inner_end_y':17,
'right_eyebrow_outer_end_x':18,
'right_eyebrow_outer_end_y':19,
'nose_tip_x':20,
'nose_tip_y':21,
'mouth_left_corner_x':22,
'mouth_left_corner_y':23,
'mouth_right_corner_x':24,
'mouth_right_corner_y':25,
'mouth_center_top_lip_x':26,
'mouth_center_top_lip_y':27,
'mouth_center_bottom_lip_x':28,
'mouth_center_bottom_lip_y':29
}
X,y = input_data()
print(X)
print(y)
X_test,y_test = input_data(test=True)
print(X_test)
print(y_test)
y_pred = []
#
# output_file = open('submit.csv','w')
# output_file.write('RowId,Location\n')
#
# IdLookupTable = open('IdLookupTable.csv')
# IdLookupTable.readline()
#
# for line in IdLookupTable:
# RowId,ImageId,FeatureName = line.rstrip().split(',')
# image_index = int(ImageId) - 1
# feature_index = keypoint_index[FeatureName]
# feature_location = y_pred[image_index][feature_index] * 96
# output_file.write('{0},{1}\n'.format(RowId,feature_location))
#
# output_file.close()
# IdLookupTable.close()
model = Sequential()
model.add(Conv2D(32, (3, 3), padding='same',
input_shape=X.shape[1:]))
model.add(Activation('relu'))
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(64, (3, 3), padding='same'))
model.add(Activation('relu'))
model.add(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(512))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(30))
# initiate RMSprop optimizer
opt = keras.optimizers.rmsprop(lr=0.0001, decay=1e-6)
# Let's train the model using RMSprop
model.compile(loss='mse',
optimizer=opt,
metrics=['accuracy'])
x_train = X.astype('float32')
x_test = X_test.astype('float32')
# x_train /= 255
# x_test /= 255
history = model.fit(x_train, y,batch_size=BATCH_SIZE,epochs=EPOCHS,validation_split=0.2,shuffle=True)
pyplot.plot(history.history['loss'], label='train')
pyplot.plot(history.history['val_loss'], label='test')
pyplot.title('model train vs validation loss')
pyplot.ylabel('loss')
pyplot.xlabel('epoch')
pyplot.legend(['train_loss','test_loss','tran_acc','test_acc'],loc='upper right')
pyplot.legend(['train_loss','test_loss'],loc='upper right')
pyplot.show()
model.save("keypointdetection")
model = load_model("keypointdetection")
y_test = model.predict(x_test)
print(y_test)
y_pred = y_test
def plot_sample(x, y, axis):
img = x.reshape(96, 96)
axis.imshow(img, cmap='gray')
axis.scatter(y[0::2] *96, y[1::2] *96, marker='x', s=10)
fig = pyplot.figure(figsize=(6, 6))
fig.subplots_adjust(
left=0, right=1, bottom=0, top=1, hspace=0.05, wspace=0.05)
for iin range(16):
ax = fig.add_subplot(4, 4, i +1, xticks=[], yticks=[])
plot_sample(X_test[i], y_pred[i], ax)
pyplot.show()
第四部分：测试卷积神经网络的性能
我们刚刚训练的model对象已经保存了训练时打印在控制台桌面中的记录，我们可以获取这个记录通过history 相关属性，让我们画出这两个曲线。
pyplot.plot(history.history['loss'], label='train')
pyplot.plot(history.history['val_loss'], label='test')
pyplot.title('model train vs validation loss')
pyplot.ylabel('loss')
pyplot.xlabel('epoch')
pyplot.legend(['train_loss','test_loss','tran_acc','test_acc'],loc='upper right')
pyplot.legend(['train_loss','test_loss'],loc='upper right')
pyplot.show()
y_test = model.predict(x_test)
print(y_test)
y_pred = y_test
def plot_sample(x, y, axis):
img = x.reshape(96, 96)
axis.imshow(img, cmap='gray')
axis.scatter(y[0::2] *96, y[1::2] *96, marker='x', s=10)
fig = pyplot.figure(figsize=(6, 6))
fig.subplots_adjust(
left=0, right=1, bottom=0, top=1, hspace=0.05, wspace=0.05)
for iin range(16):
ax = fig.add_subplot(4, 4, i +1, xticks=[], yticks=[])
plot_sample(X_test[i], y_pred[i], ax)
pyplot.show()
卷积神经网络预测的结果（从测试集抽出了16个样例）
预测结果看起来还不错，但是有点时候还是有一点偏。让我们试着做的更好一些。