一般来说,用CNN来进行图像分类任务可以获得很高的准确率,而LSTM一般用于自然语言处理方面,但是,用LSTM也是可以实现图像分类并得到较高准确率的。代码如下:
import tensorflow as tf import tensorflow.keras as keras from tensorflow.keras
import models, layers, optimizers import matplotlib.pyplot as plt # Mnist数据集加载 (
x_train_all, y_train_all), (x_test, y_test) = keras.datasets.mnist.load_data()
# Mnist数据集简单归一化 x_train_all, x_test = x_train_all / 255.0, x_test / 255.0
x_train, x_test = x_train_all[:50000], x_train_all[50000:] y_train, y_test =
y_train_all[:50000], y_train_all[50000:] print(x_train.shape) # 构建模型 inputs =
layers.Input(shape=(x_train.shape[1], x_train.shape[2]), name='inputs') print(
inputs.shape) lstm = layers.LSTM(units=128, return_sequences=False)(inputs)
print(lstm.shape) outputs = layers.Dense(10, activation='softmax')(lstm) print(
outputs.shape) lstm = keras.Model(inputs, outputs) # 查看模型 lstm.compile(optimizer
=keras.optimizers.Adam(0.001), loss='sparse_categorical_crossentropy', metrics=[
'accuracy']) lstm.summary() # 训练模型 history = lstm.fit(x_train, y_train,
batch_size=32, epochs=50, validation_split=0.1) # 绘制准确率图像 data = {} data[
'accuracy'] = history.history['accuracy'] data['val_accuracy'] = history.history
['val_accuracy'] pd.DataFrame(data).plot(figsize=(8, 5)) plt.grid(True) plt.axis
([0, 30, 0, 1]) plt.show() # 绘制损失图像 data = {} data['loss'] = history.history[
'loss'] data['val_loss'] = history.history['val_loss'] pd.DataFrame(data).plot(
figsize=(8, 5)) plt.grid(True) plt.axis([0, 30, 0, 1]) plt.show() # 验证模型: lstm.
evaluate(x_test, y_test, verbose=2)

技术
今日推荐
阅读数 169811
下载桌面版
GitHub
百度网盘(提取码:draw)
Gitee
云服务器优惠
阿里云优惠券
腾讯云优惠券
华为云优惠券
站点信息
问题反馈
邮箱:ixiaoyang8@qq.com
QQ群:766591547
关注微信