如何导入预训练的 InceptionV4 模型以在 Kaggle 中训练我们的模型？

Question

inputs = Input(shape=(img_dims, img_dims, 3))
from keras.applications.inception_v3 import InceptionV3
base_model = InceptionV3(weights='imagenet',include_top=False,input_shape=(img_dims, img_dims, 3))
x = base_model.output
x = Dropout(0.5)(x)
from keras.layers import GlobalAveragePooling2D
x = GlobalAveragePooling2D()(x)
x = Dense(128,activation='relu')(x)
x = BatchNormalization()(x)
output = Dense(1,activation = 'sigmoid')(x)

How to replace inception v3 with inception v4 and call the required weights?

Answer 1

Currently to my knowledge there is no API available to use InceptionV4 in Keras.

Instead, you can create the InceptionV4 network and load the pretrained weights in the created network in this link .

To create InceptionV4 and use it in your code, you can refer the link here.