输出利用resnet读取自己搭建cifar10文件提取输出图像特征的python代码

以下是一个示例代码，可以读取自己搭建的CIFAR10数据集，使用ResNet模型提取图像特征。

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras.datasets import cifar10

# 加载CIFAR10数据集
(x_train, y_train), (x_test, y_test) = cifar10.load_data()

# 数据预处理
x_train = x_train.astype("float32") / 255.0
x_test = x_test.astype("float32") / 255.0

# 定义ResNet模型
def resnet_block(inputs, num_filters, kernel_size, strides, activation="relu"):
    x = layers.Conv2D(num_filters, kernel_size=kernel_size, strides=strides, padding="same")(inputs)
    x = layers.BatchNormalization()(x)
    if activation:
        x = layers.Activation(activation)(x)
    return x

def resnet18():
    inputs = keras.Input(shape=(32, 32, 3))
    x = resnet_block(inputs, num_filters=64, kernel_size=3, strides=1)
    x = resnet_block(x, num_filters=64, kernel_size=3, strides=1)
    x = resnet_block(x, num_filters=128, kernel_size=3, strides=2)
    x = resnet_block(x, num_filters=128, kernel_size=3, strides=1)
    x = resnet_block(x, num_filters=256, kernel_size=3, strides=2)
    x = resnet_block(x, num_filters=256, kernel_size=3, strides=1)
    x = resnet_block(x, num_filters=512, kernel_size=3, strides=2)
    x = resnet_block(x, num_filters=512, kernel_size=3, strides=1)
    x = layers.GlobalAveragePooling2D()(x)
    outputs = layers.Dense(10, activation="softmax")(x)
    return keras.Model(inputs, outputs)

model = resnet18()

# 编译模型
model.compile(
    loss="sparse_categorical_crossentropy",
    optimizer=keras.optimizers.Adam(lr=0.001),
    metrics=["accuracy"],
)

# 训练模型
model.fit(x_train, y_train, batch_size=64, epochs=10, validation_split=0.2)

# 提取图像特征
feat_extractor = keras.Model(inputs=model.inputs, outputs=model.layers[-2].output)
features = feat_extractor(x_train)
print(features.shape)