0
我使用的下一个keras基于架构(article):为什么使用keras Conv2D图层时会出现错误?检查时出现错误:
model = Sequential()
model.add(Conv2D(32, (3, 3), input_shape=(150, 150, 3)))
...
model.fit_generator(
train_generator,
steps_per_epoch=nb_train_samples // batch_size,
epochs=epochs,
validation_data=validation_generator,
validation_steps=nb_validation_samples // batch_size)
model.save_weights('first_try.h5')
from keras.preprocessing.image import ImageDataGenerator, array_to_img, img_to_array, load_img
img = load_img('test_data/a1.jpg') # this is a PIL image
img = img.resize((150, 150))
x = img_to_array(img)
prediction = model.predict(x)
print(prediction)
ValueError异常预期conv2d_1_input有4个维度,但得到了阵列形状(150,150,3)
灿你请告诉我如何解决它?
还我发现X = np.expand_dims(X,轴= 0) –
@OlegDats准确,甚至一个简单的'X .reshape(1,* img.shape)'应该可以工作。 – 5agado