我正在使用tensorflow 2.2.0rc,并注意到sudo apt-get install git
内部的InceptionV3的实现与其实现的论文inception paper之间存在一些差异。
tf.keras.applications.inception_v3.InceptionV3
branch1x1 = conv2d_bn(x, 64, 1, 1)
# Shouldnt this be
# branch3x3 = conv2d_bn(x, 48, 1, 1)
# branch3x3 = conv2d_bn(branch3x3, 64, 3, 3)
branch5x5 = conv2d_bn(x, 48, 1, 1)
branch5x5 = conv2d_bn(branch5x5, 64, 5, 5)
branch3x3dbl = conv2d_bn(x, 64, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch_pool = layers.AveragePooling2D(
(3, 3), strides=(1, 1), padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 32, 1, 1)
x = layers.concatenate([branch1x1, branch5x5, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed0')
真的不是branch5x5
吗?