model.predict() - 精度接近 1 的模型预测错误类

Question

我创建了一个模型来识别车牌。就是这个：

def create_model(input_shape = (224, 224, 3)):
    input_img = Input(shape=input_shape)
    model = efnB0_model (input_img)
    model = GlobalAveragePooling2D(name='avg_pool')(model)
    model = Dropout(0.2)(model)
    backbone = model

    branches = []
    for i in range(7):
            branches.append(backbone)
            branches[i] = Dense(360, name="branch_"+str(i)+"_Dense_360")(branches[i])
            branches[i] = BatchNormalization()(branches[i])
            branches[i] = Activation("relu") (branches[i])
            branches[i] = Dropout(0.2)(branches[i])
                       
            branches[i] = Dense(35, activation = "softmax", name="branch_"+str(i)+"_output")(branches[i])
        
    output = Concatenate(axis=1)(branches)
    output = Reshape((7, 35))(output)
    model = Model(input_img, output)

    return model

我使用了这个 DataGenerator：

import tensorflow.keras as keras
from skimage.io import imread
from skimage.transform import resize
import numpy as np
import math

class DataGenerator(Sequence):

    def __init__(self, x_set, y_set, batch_size):
        self.x, self.y = x_set, y_set
        self.batch_size = batch_size

    def __len__(self):
        return math.ceil(len(self.x) / self.batch_size)

    def __getitem__(self, idx):
        batch_x = self.x[idx*self.batch_size : (idx + 1)*self.batch_size]
        batch_x = np.array([resize(imread(file_name), (224, 224)) for file_name in batch_x])
        batch_x = batch_x * 1./255
        batch_y = self.y[idx*self.batch_size : (idx + 1)*self.batch_size]
        batch_y = np.array(batch_y)

        return batch_x, batch_y

因此，我使用以下代码对每个车牌进行了热编码（长度为 7 个字符，每个位置可能有 35 个字符：

#One Hot Encoding der Labels, Zielarray hat eine Shape von (7,35)
from numpy import argmax
# define input string

def my_onehot_encoded(label):
    # define universe of possible input values
    characters = '0123456789ABCDEFGHIJKLMNPQRSTUVWXYZ'
    # define a mapping of chars to integers
    char_to_int = dict((c, i) for i, c in enumerate(characters))
    int_to_char = dict((i, c) for i, c in enumerate(characters))
    # integer encode input data
    integer_encoded = [char_to_int[char] for char in label]
    # one hot encode
    onehot_encoded = list()
    for value in integer_encoded:
        character = [0 for _ in range(len(characters))]
        character[value] = 1
        onehot_encoded.append(character)

    return onehot_encoded

对于标签为“7CT2498”的许可证，我得到以下 onehot 编码输出：

[[0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0], [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]

现在，当运行我在上面为 10.000 个训练数据和 3.000 个验证数据创建的模型时，我得到的训练精度为 0.9969，验证精度为 0.9798，所以还不错。

但现在我尝试用这个模型预测车牌（图像来自与我的训练和验证数据相同的数据集）。

我使用了这个代码：

model = keras.models.load_model(
    "/path/to/model.h5", compile=True)
opt = keras.optimizers.Adam(learning_rate=0.0001)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=["accuracy"])
img = cv2.imread('/path/to/image.png')
img = cv2.resize(img,(224,224))
img = np.reshape(img,[1,224,224,3])

classes = model.predict(img)

print(classes)

而且我只正确预测了一个 class。我的代码有问题吗？

Answer 1

看起来你的模型过度拟合了很多......你可能想研究一下这个问题。另一点，对于识别车牌你可能想选择不同的方法，你可以使用这些： -

OpenCV 平板检测器
图像分类器
（不推荐）Pytesseract

此致，无限S

Answer 2

查看您发给我的代码后，您似乎在拟合模型时使用 skimage 进行预处理，在进行预测时使用 opencv。使用相同的预处理代码后，它工作正常：

from skimage.io import imread
from skimage.transform import resize
import numpy as np
import math

img = imread('path/to/image')
img = resize(img,(224,224))
img = img*1./255
img = np.reshape(img,[1,224,224,3])

classes = model.predict(img)

print(classes)

model.predict() - 精度接近 1 的模型预测错误类

model.predict() - Model with accuracy near to 1 predicts wrong classes

python

predict

conv-neural-network

tensorflow

model.predict() - 精度接近 1 的模型预测错误 类

model.predict() - Model with accuracy near to 1 predicts wrong classes

python

predict

conv-neural-network

tensorflow

model.predict() - 精度接近 1 的模型预测错误类