ValueError：数据基数不明确：x 大小：12000 y 大小：640 确保所有数组包含相同数量的样本

Question

我尝试使用 tensorflow 训练数据集，但数据基数出现错误。

*代码

import numpy as np
import os
from PIL import Image
import matplotlib.pyplot as plt
import cv2
from sklearn.model_selection import train_test_split
import tensorflow as tf

def read_patches(Image_path, label):
    patches = []
    labels = []
    Image_File_name = []

    for (path,dir,files) in os.walk(Image_path):
        files.sort()
        for filename in files :
            if('good' in filename):
                Image_File_name.append(Image_path+'\\'+filename)


    for i in range(len(Image_File_name)):
        # print(Image_File_name[i])
        patch = cv2.imread(Image_File_name[i])
        gray = cv2.cvtColor(patch, cv2.COLOR_BGR2GRAY) #RGB to Gray

        patches.append(gray)
        labels.append(label)

    patches = np.array(patches)
    labels = np.array(labels)

    patchDim = patches.shape
    patches = np.reshape(patches, (patchDim[0], patchDim[1], patchDim[2], 1) )
    labels = np.reshape(labels, (len(labels), 1) )

    return patches, labels

X, Y = read_patches("C:\\works\\data\\image_source_01\\good", 1) 

train_X = train_X.reshape(-1, 32, 32, 1)
test_X = test_X.reshape(-1, 32, 32, 1)
print(train_X.shape, test_X.shape)

model = tf.keras.Sequential([
      tf.keras.layers.Conv2D(input_shape=(32,32,1), kernel_size=(3, 3), filters = 32, 
padding='same', activation='relu'),
      tf.keras.layers.Conv2D(kernel_size=(3, 3), filters = 64, padding='same', activation='relu'),
      tf.keras.layers.MaxPool2D(strides=(2, 2)),
      tf.keras.layers.Dropout(rate=0.5),
      tf.keras.layers.Conv2D(kernel_size=(3, 3), filters = 128, padding='same', activation='relu'),
      tf.keras.layers.Conv2D(kernel_size=(3, 3), filters = 256, padding='valid', activation='relu'),
      tf.keras.layers.MaxPool2D(strides=(2, 2)),
      tf.keras.layers.Dropout(rate=0.5),
      tf.keras.layers.Flatten(),
      tf.keras.layers.Dense(units=512, activation='relu'),
      tf.keras.layers.Dropout(rate=0.5),
      tf.keras.layers.Dense(units=256, activation='relu'),
      tf.keras.layers.Dropout(rate=0.5),
      tf.keras.layers.Dense(units=10, activation='softmax')                        
])

model.compile(optimizer=tf.keras.optimizers.Adam(),
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])
          
model.summary()

history = model.fit(train_X, train_Y, epochs=25, validation_split=0.25)

*错误

Traceback（最近一次调用最后一次）： 文件“C:\works\test4.py”，第 139 行，在 历史 = model.fit(train_X, train_Y, epochs=25, validation_split=0.25) ... packages\keras\engine\data_adapter.py”，第 1649 行，在 _check_data_cardinality 引发 ValueError（味精） ValueError：数据基数不明确： x 尺寸：12000 尺码：640 确保所有数组都包含相同数量的样本。

Answer 1

您的错误说明了一切：您需要确保 train_X 和 train_Y 具有相同的形状。如果您像这样加载 mnist 数据，您会注意到 x_train 的形状为 (60000, 28, 28, 1) 和 y_train 的形状为 (60000,)：

(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()

请注意，两者都有 60000 个样本。您的 train_X 有 12000 个图像，train_Y 有 640 个标签，这是行不通的，因为每个图像都需要一个相应的标签。此外，我认为您不必重塑数据，但这取决于它在此之前的样子。