Question

刚开始使用TensorFlow和Keras建立我的第一个模型。该模型的评估精度达到98％，但在预测模型时遇到困难。

import tensorflow as tf
import numpy as np
from PIL import Image
from keras.preprocessing import image

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()

tf.keras.backend.image_data_format()== 'channels_last'


x_train, x_test = x_train / 255.0, x_test / 255.0

model = tf.keras.models.Sequential([
  tf.keras.layers.Flatten(),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dropout(0.2),
  tf.keras.layers.Dense(10, activation='softmax')
])

print(x_train.shape)

model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])


model.fit(x_train, y_train, epochs=5)

model.evaluate(x_test,  y_test, verbose=2)

img = Image.open('8.png')
img = img.resize((28,28)
im2arr = np.array(img)
im2arr = im2arr.reshape(1,28,28,1)
print(im2arr.shape)
prediction = model.predict_classes(im2arr)
print(prediction)

Answer 1

我不知道您的8.png里面有什么，但是我更改了代码中的一些内容，对我来说，它可以处理一些测试图像。

import tensorflow as tf
import numpy as np
import PIL
from PIL import Image
from keras.preprocessing import image


mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()

tf.keras.backend.image_data_format()== 'channels_last'


x_train, x_test = x_train / 255.0, x_test / 255.0

model = tf.keras.models.Sequential([
  tf.keras.layers.Flatten(),
  tf.keras.layers.GaussianNoise(0.05),
  tf.keras.layers.Dense(32, activation='relu'),
  tf.keras.layers.Dropout(0.2),
  tf.keras.layers.Dense(10, activation='softmax')
])

print(x_train.shape)

model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])


model.fit(x_train, y_train, epochs=10)

model.evaluate(x_test,  y_test, verbose=2)

img = Image.open('8.png').convert('L')
img = img.resize((28, 28), resample=PIL.Image.BILINEAR)
im2arr = 255. - np.array(img, dtype=np.float32)
im2arr = im2arr / 255.
im2arr = im2arr.reshape(1,28,28,1)
print(im2arr.shape)
prediction = model.predict_classes(im2arr)
print(prediction)

基本上，我只是修改了模型以使其对过度拟合更为健壮，并在将输入图像提供给模型之前为输入图像添加了一些准备。

为什么我的keras模型无法获得正确的结果

1 个答案: