参考论文：Rethinking the Inception Architecture for Computer Vision作者：Christian Szegedy; Vincent Vanhoucke; Sergey Ioffe; Jon Shlens; Zbigniew Wojna
1、Inception-v2/v3优化策略
Inception-v2和Inception-v3都是出自同一篇论文《Rethinking the inception architecture for computer vision》，该论文提出了多种基于 Inception-v1 的模型优化 方法，Inception-v2 用了其中的一部分模型优化方法，Inception-v3 用了论文中提到的所有 优化方法。相当于 Inception-v2 只是一个过渡版本，Inception-v3 一般用得更多。
Inception-v3 最大的优化是模型结构上的优化，在 Inception-v3 中作者对 Inception 结 构中的卷积进行了分解。分解后的好处是增加了网络的层数，也就是增加了网络的特征提取能力。同时作者还对 Inception 结构进行了一些调整，设计了不同的 Inception，用在模型 的不同位置。
先回忆一下最原始的Inception结构，如下图所示
图中的 convolutions 表示卷积，filter concatenation 表示滤波器合并，max pooling 表示最大池化，previous layer 表示前一层。
Inception-v3 中提出了一个新思路，可以使用两个 3×3 卷积来替代原始 Inception 结构 中的 5×5 卷积，如下图所示。
将 5×5 卷积分解为两层的 3×3 卷积，对于最后得到的特征来说，感受野的大小是相同 的，都是 5×5 的区域。相当于 5×5 卷积对 5×5 区域进行特征提取，得到一个特征值；两层的 3×3 卷积对 5×5 区域进行特征提取，也是得到一个特征值。这两种特征提取的方式类 似，不过最后得到的特征值可能是不同的，两层 3×3 卷积做了两次卷积得到的特征值或许会更好一些。
沿着这个卷积分解的思路继续思考，作者又提出了一种新的卷积分解，把 3×3 卷积分解 为 1×3 卷积和 3×1 卷积，如下图所示。
把 3×3 卷积分解为 1×3 卷积和 3×1 卷积，道理跟将 5×5 卷积分解为两层的 3×3 卷积 差不多，对于最后的特征来说，感受野的大小是一样的，并且分解后可以让网络层数变得更 多，增加网络的非线性。理论上 n×n 的卷积都可以分解为 1×n 卷积和 n×1 卷积。
作者还分析了减小特征图大小时的操作，如下图所示。
作者认为直接使用窗口大小 2×2，步长为 2 的池化来压缩特征图的大小效果不太好。因 为特征图的数量不变，但是特征图的长宽变成为原来的 1/2，相当于特征值的数量被压缩为 原来的 1/4 了，特征值的数量一下减少太多不利于模型的训练，所以左边的结构不太理想。 右边的结构先用 Inception 来增加特征图数量然后再进行池化减小特征图大小，对于特征的 提取来说没什么问题，就是计算量太大。
所以设计了新的 Inception 结构，在减小特征图大小的同时可以增加特征图的数量，如下图所示。
图中 Filter Concat 表示滤波器拼接；stride 表示步长；concat 表示拼接；conv 表示卷 积；pool 表示池化。
除此之外作者还根据实验分析和建模经验，设计了一些新的 Inception 结构，如下图所示。
2、Inception-v2/v3模型结构
Inception-v2/v3 模型的结构非常庞大，Inception-v2/v3 论文中给出的模型结构描述也 不是特别清晰，结构如图所示。
Filter Concat 表示滤波器拼接；Pool 表示池化；type 表 示层的类型；patch size/stride or remarks 表示窗口大小/步长；input size 表示输入大 小；conv 表示卷积；pool 表示池化；linear 表示全连接层。
figure5:将5*5卷积分解为2个3*3卷积
figure6:分解为不对称卷积
figure7:扩展滤波器组
3、Inception-v3在ImageNet数据集中的表现
Inception-v3 单模型测试结果：
图中 Network 表示网络；Crops Evaluated 表示模型评估时裁剪出多少张图片进行预 测；Top-5 Error 表示 Top5 错误率；Top-1 Error 表示 Top1 错误率。
Inception-v3 模型融合后的测试结果。
图中 Network 表示网络；Models Evaluated 表示评估时集成了几个模型；Crops Evaluated 表示模型评估时裁剪出多少张图片进行预测；Top-5 Error 表示 Top5 错误率； Top-1 Error 表示 Top1 错误率。
Inception-v3 模型融合后的 Top5 错误率为 3.58%，这个结果跟 2015 年 ImageNet Challenge 图像识别比赛的冠军 ResNet 已经非常接近，ResNet 的 Top5 错误率为 3.57%。
4、迁移学习复现网络结构
import tensorflow as tf
from tensorflow.keras.layers import *
from tensorflow.keras.applications import InceptionV3
from plot_model import plot_model
model=InceptionV3(
input_shape=(299,299,3),
include_top=True, # 是否包含最后的全连接层
weights='imagenet', # 权重默认为ImageNet一千类图像分类
classes=1000 # 1000个类别
)
model.summary()
plot_model(model,to_file='img/inception-v3.png',show_shapes=True)Model: "inception_v3"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_1 (InputLayer) [(None, 299, 299, 3) 0
__________________________________________________________________________________________________
conv2d (Conv2D) (None, 149, 149, 32) 864 input_1[0][0]
__________________________________________________________________________________________________
batch_normalization (BatchNorma (None, 149, 149, 32) 96 conv2d[0][0]
__________________________________________________________________________________________________
activation (Activation) (None, 149, 149, 32) 0 batch_normalization[0][0]
__________________________________________________________________________________________________
conv2d_1 (Conv2D) (None, 147, 147, 32) 9216 activation[0][0]
__________________________________________________________________________________________________
batch_normalization_1 (BatchNor (None, 147, 147, 32) 96 conv2d_1[0][0]
__________________________________________________________________________________________________
activation_1 (Activation) (None, 147, 147, 32) 0 batch_normalization_1[0][0]
__________________________________________________________________________________________________
conv2d_2 (Conv2D) (None, 147, 147, 64) 18432 activation_1[0][0]
__________________________________________________________________________________________________
batch_normalization_2 (BatchNor (None, 147, 147, 64) 192 conv2d_2[0][0]
__________________________________________________________________________________________________
activation_2 (Activation) (None, 147, 147, 64) 0 batch_normalization_2[0][0]
__________________________________________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 73, 73, 64) 0 activation_2[0][0]
__________________________________________________________________________________________________
conv2d_3 (Conv2D) (None, 73, 73, 80) 5120 max_pooling2d[0][0]
__________________________________________________________________________________________________
batch_normalization_3 (BatchNor (None, 73, 73, 80) 240 conv2d_3[0][0]
__________________________________________________________________________________________________
activation_3 (Activation) (None, 73, 73, 80) 0 batch_normalization_3[0][0]
__________________________________________________________________________________________________
conv2d_4 (Conv2D) (None, 71, 71, 192) 138240 activation_3[0][0]
__________________________________________________________________________________________________
batch_normalization_4 (BatchNor (None, 71, 71, 192) 576 conv2d_4[0][0]
__________________________________________________________________________________________________
activation_4 (Activation) (None, 71, 71, 192) 0 batch_normalization_4[0][0]
__________________________________________________________________________________________________
max_pooling2d_1 (MaxPooling2D) (None, 35, 35, 192) 0 activation_4[0][0]
__________________________________________________________________________________________________
conv2d_8 (Conv2D) (None, 35, 35, 64) 12288 max_pooling2d_1[0][0]
__________________________________________________________________________________________________
batch_normalization_8 (BatchNor (None, 35, 35, 64) 192 conv2d_8[0][0]
__________________________________________________________________________________________________
activation_8 (Activation) (None, 35, 35, 64) 0 batch_normalization_8[0][0]
__________________________________________________________________________________________________
conv2d_6 (Conv2D) (None, 35, 35, 48) 9216 max_pooling2d_1[0][0]
__________________________________________________________________________________________________
conv2d_9 (Conv2D) (None, 35, 35, 96) 55296 activation_8[0][0]
__________________________________________________________________________________________________
batch_normalization_6 (BatchNor (None, 35, 35, 48) 144 conv2d_6[0][0]
__________________________________________________________________________________________________
batch_normalization_9 (BatchNor (None, 35, 35, 96) 288 conv2d_9[0][0]
__________________________________________________________________________________________________
activation_6 (Activation) (None, 35, 35, 48) 0 batch_normalization_6[0][0]
__________________________________________________________________________________________________
activation_9 (Activation) (None, 35, 35, 96) 0 batch_normalization_9[0][0]
__________________________________________________________________________________________________
average_pooling2d (AveragePooli (None, 35, 35, 192) 0 max_pooling2d_1[0][0]
__________________________________________________________________________________________________
conv2d_5 (Conv2D) (None, 35, 35, 64) 12288 max_pooling2d_1[0][0]
__________________________________________________________________________________________________
conv2d_7 (Conv2D) (None, 35, 35, 64) 76800 activation_6[0][0]
__________________________________________________________________________________________________
conv2d_10 (Conv2D) (None, 35, 35, 96) 82944 activation_9[0][0]
__________________________________________________________________________________________________
conv2d_11 (Conv2D) (None, 35, 35, 32) 6144 average_pooling2d[0][0]
__________________________________________________________________________________________________
batch_normalization_5 (BatchNor (None, 35, 35, 64) 192 conv2d_5[0][0]
__________________________________________________________________________________________________
batch_normalization_7 (BatchNor (None, 35, 35, 64) 192 conv2d_7[0][0]
__________________________________________________________________________________________________
batch_normalization_10 (BatchNo (None, 35, 35, 96) 288 conv2d_10[0][0]
__________________________________________________________________________________________________
batch_normalization_11 (BatchNo (None, 35, 35, 32) 96 conv2d_11[0][0]
__________________________________________________________________________________________________
activation_5 (Activation) (None, 35, 35, 64) 0 batch_normalization_5[0][0]
__________________________________________________________________________________________________
activation_7 (Activation) (None, 35, 35, 64) 0 batch_normalization_7[0][0]
__________________________________________________________________________________________________
activation_10 (Activation) (None, 35, 35, 96) 0 batch_normalization_10[0][0]
__________________________________________________________________________________________________
activation_11 (Activation) (None, 35, 35, 32) 0 batch_normalization_11[0][0]
__________________________________________________________________________________________________
mixed0 (Concatenate) (None, 35, 35, 256) 0 activation_5[0][0]
activation_7[0][0]
activation_10[0][0]
activation_11[0][0]
__________________________________________________________________________________________________
conv2d_15 (Conv2D) (None, 35, 35, 64) 16384 mixed0[0][0]
__________________________________________________________________________________________________
batch_normalization_15 (BatchNo (None, 35, 35, 64) 192 conv2d_15[0][0]
__________________________________________________________________________________________________
activation_15 (Activation) (None, 35, 35, 64) 0 batch_normalization_15[0][0]
__________________________________________________________________________________________________
conv2d_13 (Conv2D) (None, 35, 35, 48) 12288 mixed0[0][0]
__________________________________________________________________________________________________
conv2d_16 (Conv2D) (None, 35, 35, 96) 55296 activation_15[0][0]
__________________________________________________________________________________________________
batch_normalization_13 (BatchNo (None, 35, 35, 48) 144 conv2d_13[0][0]
__________________________________________________________________________________________________
batch_normalization_16 (BatchNo (None, 35, 35, 96) 288 conv2d_16[0][0]
__________________________________________________________________________________________________
activation_13 (Activation) (None, 35, 35, 48) 0 batch_normalization_13[0][0]
__________________________________________________________________________________________________
activation_16 (Activation) (None, 35, 35, 96) 0 batch_normalization_16[0][0]
__________________________________________________________________________________________________
average_pooling2d_1 (AveragePoo (None, 35, 35, 256) 0 mixed0[0][0]
__________________________________________________________________________________________________
conv2d_12 (Conv2D) (None, 35, 35, 64) 16384 mixed0[0][0]
__________________________________________________________________________________________________
conv2d_14 (Conv2D) (None, 35, 35, 64) 76800 activation_13[0][0]
__________________________________________________________________________________________________
conv2d_17 (Conv2D) (None, 35, 35, 96) 82944 activation_16[0][0]
__________________________________________________________________________________________________
conv2d_18 (Conv2D) (None, 35, 35, 64) 16384 average_pooling2d_1[0][0]
__________________________________________________________________________________________________
batch_normalization_12 (BatchNo (None, 35, 35, 64) 192 conv2d_12[0][0]
__________________________________________________________________________________________________
batch_normalization_14 (BatchNo (None, 35, 35, 64) 192 conv2d_14[0][0]
__________________________________________________________________________________________________
batch_normalization_17 (BatchNo (None, 35, 35, 96) 288 conv2d_17[0][0]
__________________________________________________________________________________________________
batch_normalization_18 (BatchNo (None, 35, 35, 64) 192 conv2d_18[0][0]
__________________________________________________________________________________________________
activation_12 (Activation) (None, 35, 35, 64) 0 batch_normalization_12[0][0]
__________________________________________________________________________________________________
activation_14 (Activation) (None, 35, 35, 64) 0 batch_normalization_14[0][0]
__________________________________________________________________________________________________
activation_17 (Activation) (None, 35, 35, 96) 0 batch_normalization_17[0][0]
__________________________________________________________________________________________________
activation_18 (Activation) (None, 35, 35, 64) 0 batch_normalization_18[0][0]
__________________________________________________________________________________________________
mixed1 (Concatenate) (None, 35, 35, 288) 0 activation_12[0][0]
activation_14[0][0]
activation_17[0][0]
activation_18[0][0]
__________________________________________________________________________________________________
conv2d_22 (Conv2D) (None, 35, 35, 64) 18432 mixed1[0][0]
__________________________________________________________________________________________________
batch_normalization_22 (BatchNo (None, 35, 35, 64) 192 conv2d_22[0][0]
__________________________________________________________________________________________________
activation_22 (Activation) (None, 35, 35, 64) 0 batch_normalization_22[0][0]
__________________________________________________________________________________________________
conv2d_20 (Conv2D) (None, 35, 35, 48) 13824 mixed1[0][0]
__________________________________________________________________________________________________
conv2d_23 (Conv2D) (None, 35, 35, 96) 55296 activation_22[0][0]
__________________________________________________________________________________________________
batch_normalization_20 (BatchNo (None, 35, 35, 48) 144 conv2d_20[0][0]
__________________________________________________________________________________________________
batch_normalization_23 (BatchNo (None, 35, 35, 96) 288 conv2d_23[0][0]
__________________________________________________________________________________________________
activation_20 (Activation) (None, 35, 35, 48) 0 batch_normalization_20[0][0]
__________________________________________________________________________________________________
activation_23 (Activation) (None, 35, 35, 96) 0 batch_normalization_23[0][0]
__________________________________________________________________________________________________
average_pooling2d_2 (AveragePoo (None, 35, 35, 288) 0 mixed1[0][0]
__________________________________________________________________________________________________
conv2d_19 (Conv2D) (None, 35, 35, 64) 18432 mixed1[0][0]
__________________________________________________________________________________________________
conv2d_21 (Conv2D) (None, 35, 35, 64) 76800 activation_20[0][0]
__________________________________________________________________________________________________
conv2d_24 (Conv2D) (None, 35, 35, 96) 82944 activation_23[0][0]
__________________________________________________________________________________________________
conv2d_25 (Conv2D) (None, 35, 35, 64) 18432 average_pooling2d_2[0][0]
__________________________________________________________________________________________________
batch_normalization_19 (BatchNo (None, 35, 35, 64) 192 conv2d_19[0][0]
__________________________________________________________________________________________________
batch_normalization_21 (BatchNo (None, 35, 35, 64) 192 conv2d_21[0][0]
__________________________________________________________________________________________________
batch_normalization_24 (BatchNo (None, 35, 35, 96) 288 conv2d_24[0][0]
__________________________________________________________________________________________________
batch_normalization_25 (BatchNo (None, 35, 35, 64) 192 conv2d_25[0][0]
__________________________________________________________________________________________________
activation_19 (Activation) (None, 35, 35, 64) 0 batch_normalization_19[0][0]
__________________________________________________________________________________________________
activation_21 (Activation) (None, 35, 35, 64) 0 batch_normalization_21[0][0]
__________________________________________________________________________________________________
activation_24 (Activation) (None, 35, 35, 96) 0 batch_normalization_24[0][0]
__________________________________________________________________________________________________
activation_25 (Activation) (None, 35, 35, 64) 0 batch_normalization_25[0][0]
__________________________________________________________________________________________________
mixed2 (Concatenate) (None, 35, 35, 288) 0 activation_19[0][0]
activation_21[0][0]
activation_24[0][0]
activation_25[0][0]
__________________________________________________________________________________________________
conv2d_27 (Conv2D) (None, 35, 35, 64) 18432 mixed2[0][0]
__________________________________________________________________________________________________
batch_normalization_27 (BatchNo (None, 35, 35, 64) 192 conv2d_27[0][0]
__________________________________________________________________________________________________
activation_27 (Activation) (None, 35, 35, 64) 0 batch_normalization_27[0][0]
__________________________________________________________________________________________________
conv2d_28 (Conv2D) (None, 35, 35, 96) 55296 activation_27[0][0]
__________________________________________________________________________________________________
batch_normalization_28 (BatchNo (None, 35, 35, 96) 288 conv2d_28[0][0]
__________________________________________________________________________________________________
activation_28 (Activation) (None, 35, 35, 96) 0 batch_normalization_28[0][0]
__________________________________________________________________________________________________
conv2d_26 (Conv2D) (None, 17, 17, 384) 995328 mixed2[0][0]
__________________________________________________________________________________________________
conv2d_29 (Conv2D) (None, 17, 17, 96) 82944 activation_28[0][0]
__________________________________________________________________________________________________
batch_normalization_26 (BatchNo (None, 17, 17, 384) 1152 conv2d_26[0][0]
__________________________________________________________________________________________________
batch_normalization_29 (BatchNo (None, 17, 17, 96) 288 conv2d_29[0][0]
__________________________________________________________________________________________________
activation_26 (Activation) (None, 17, 17, 384) 0 batch_normalization_26[0][0]
__________________________________________________________________________________________________
activation_29 (Activation) (None, 17, 17, 96) 0 batch_normalization_29[0][0]
__________________________________________________________________________________________________
max_pooling2d_2 (MaxPooling2D) (None, 17, 17, 288) 0 mixed2[0][0]
__________________________________________________________________________________________________
mixed3 (Concatenate) (None, 17, 17, 768) 0 activation_26[0][0]
activation_29[0][0]
max_pooling2d_2[0][0]
__________________________________________________________________________________________________
conv2d_34 (Conv2D) (None, 17, 17, 128) 98304 mixed3[0][0]
__________________________________________________________________________________________________
batch_normalization_34 (BatchNo (None, 17, 17, 128) 384 conv2d_34[0][0]
__________________________________________________________________________________________________
activation_34 (Activation) (None, 17, 17, 128) 0 batch_normalization_34[0][0]
__________________________________________________________________________________________________
conv2d_35 (Conv2D) (None, 17, 17, 128) 114688 activation_34[0][0]
__________________________________________________________________________________________________
batch_normalization_35 (BatchNo (None, 17, 17, 128) 384 conv2d_35[0][0]
__________________________________________________________________________________________________
activation_35 (Activation) (None, 17, 17, 128) 0 batch_normalization_35[0][0]
__________________________________________________________________________________________________
conv2d_31 (Conv2D) (None, 17, 17, 128) 98304 mixed3[0][0]
__________________________________________________________________________________________________
conv2d_36 (Conv2D) (None, 17, 17, 128) 114688 activation_35[0][0]
__________________________________________________________________________________________________
batch_normalization_31 (BatchNo (None, 17, 17, 128) 384 conv2d_31[0][0]
__________________________________________________________________________________________________
batch_normalization_36 (BatchNo (None, 17, 17, 128) 384 conv2d_36[0][0]
__________________________________________________________________________________________________
activation_31 (Activation) (None, 17, 17, 128) 0 batch_normalization_31[0][0]
__________________________________________________________________________________________________
activation_36 (Activation) (None, 17, 17, 128) 0 batch_normalization_36[0][0]
__________________________________________________________________________________________________
conv2d_32 (Conv2D) (None, 17, 17, 128) 114688 activation_31[0][0]
__________________________________________________________________________________________________
conv2d_37 (Conv2D) (None, 17, 17, 128) 114688 activation_36[0][0]
__________________________________________________________________________________________________
batch_normalization_32 (BatchNo (None, 17, 17, 128) 384 conv2d_32[0][0]
__________________________________________________________________________________________________
batch_normalization_37 (BatchNo (None, 17, 17, 128) 384 conv2d_37[0][0]
__________________________________________________________________________________________________
activation_32 (Activation) (None, 17, 17, 128) 0 batch_normalization_32[0][0]
__________________________________________________________________________________________________
activation_37 (Activation) (None, 17, 17, 128) 0 batch_normalization_37[0][0]
__________________________________________________________________________________________________
average_pooling2d_3 (AveragePoo (None, 17, 17, 768) 0 mixed3[0][0]
__________________________________________________________________________________________________
conv2d_30 (Conv2D) (None, 17, 17, 192) 147456 mixed3[0][0]
__________________________________________________________________________________________________
conv2d_33 (Conv2D) (None, 17, 17, 192) 172032 activation_32[0][0]
__________________________________________________________________________________________________
conv2d_38 (Conv2D) (None, 17, 17, 192) 172032 activation_37[0][0]
__________________________________________________________________________________________________
conv2d_39 (Conv2D) (None, 17, 17, 192) 147456 average_pooling2d_3[0][0]
__________________________________________________________________________________________________
batch_normalization_30 (BatchNo (None, 17, 17, 192) 576 conv2d_30[0][0]
__________________________________________________________________________________________________
batch_normalization_33 (BatchNo (None, 17, 17, 192) 576 conv2d_33[0][0]
__________________________________________________________________________________________________
batch_normalization_38 (BatchNo (None, 17, 17, 192) 576 conv2d_38[0][0]
__________________________________________________________________________________________________
batch_normalization_39 (BatchNo (None, 17, 17, 192) 576 conv2d_39[0][0]
__________________________________________________________________________________________________
activation_30 (Activation) (None, 17, 17, 192) 0 batch_normalization_30[0][0]
__________________________________________________________________________________________________
activation_33 (Activation) (None, 17, 17, 192) 0 batch_normalization_33[0][0]
__________________________________________________________________________________________________
activation_38 (Activation) (None, 17, 17, 192) 0 batch_normalization_38[0][0]
__________________________________________________________________________________________________
activation_39 (Activation) (None, 17, 17, 192) 0 batch_normalization_39[0][0]
__________________________________________________________________________________________________
mixed4 (Concatenate) (None, 17, 17, 768) 0 activation_30[0][0]
activation_33[0][0]
activation_38[0][0]
activation_39[0][0]
__________________________________________________________________________________________________
conv2d_44 (Conv2D) (None, 17, 17, 160) 122880 mixed4[0][0]
__________________________________________________________________________________________________
batch_normalization_44 (BatchNo (None, 17, 17, 160) 480 conv2d_44[0][0]
__________________________________________________________________________________________________
activation_44 (Activation) (None, 17, 17, 160) 0 batch_normalization_44[0][0]
__________________________________________________________________________________________________
conv2d_45 (Conv2D) (None, 17, 17, 160) 179200 activation_44[0][0]
__________________________________________________________________________________________________
batch_normalization_45 (BatchNo (None, 17, 17, 160) 480 conv2d_45[0][0]
__________________________________________________________________________________________________
activation_45 (Activation) (None, 17, 17, 160) 0 batch_normalization_45[0][0]
__________________________________________________________________________________________________
conv2d_41 (Conv2D) (None, 17, 17, 160) 122880 mixed4[0][0]
__________________________________________________________________________________________________
conv2d_46 (Conv2D) (None, 17, 17, 160) 179200 activation_45[0][0]
__________________________________________________________________________________________________
batch_normalization_41 (BatchNo (None, 17, 17, 160) 480 conv2d_41[0][0]
__________________________________________________________________________________________________
batch_normalization_46 (BatchNo (None, 17, 17, 160) 480 conv2d_46[0][0]
__________________________________________________________________________________________________
activation_41 (Activation) (None, 17, 17, 160) 0 batch_normalization_41[0][0]
__________________________________________________________________________________________________
activation_46 (Activation) (None, 17, 17, 160) 0 batch_normalization_46[0][0]
__________________________________________________________________________________________________
conv2d_42 (Conv2D) (None, 17, 17, 160) 179200 activation_41[0][0]
__________________________________________________________________________________________________
conv2d_47 (Conv2D) (None, 17, 17, 160) 179200 activation_46[0][0]
__________________________________________________________________________________________________
batch_normalization_42 (BatchNo (None, 17, 17, 160) 480 conv2d_42[0][0]
__________________________________________________________________________________________________
batch_normalization_47 (BatchNo (None, 17, 17, 160) 480 conv2d_47[0][0]
__________________________________________________________________________________________________
activation_42 (Activation) (None, 17, 17, 160) 0 batch_normalization_42[0][0]
__________________________________________________________________________________________________
activation_47 (Activation) (None, 17, 17, 160) 0 batch_normalization_47[0][0]
__________________________________________________________________________________________________
average_pooling2d_4 (AveragePoo (None, 17, 17, 768) 0 mixed4[0][0]
__________________________________________________________________________________________________
conv2d_40 (Conv2D) (None, 17, 17, 192) 147456 mixed4[0][0]
__________________________________________________________________________________________________
conv2d_43 (Conv2D) (None, 17, 17, 192) 215040 activation_42[0][0]
__________________________________________________________________________________________________
conv2d_48 (Conv2D) (None, 17, 17, 192) 215040 activation_47[0][0]
__________________________________________________________________________________________________
conv2d_49 (Conv2D) (None, 17, 17, 192) 147456 average_pooling2d_4[0][0]
__________________________________________________________________________________________________
batch_normalization_40 (BatchNo (None, 17, 17, 192) 576 conv2d_40[0][0]
__________________________________________________________________________________________________
batch_normalization_43 (BatchNo (None, 17, 17, 192) 576 conv2d_43[0][0]
__________________________________________________________________________________________________
batch_normalization_48 (BatchNo (None, 17, 17, 192) 576 conv2d_48[0][0]
__________________________________________________________________________________________________
batch_normalization_49 (BatchNo (None, 17, 17, 192) 576 conv2d_49[0][0]
__________________________________________________________________________________________________
activation_40 (Activation) (None, 17, 17, 192) 0 batch_normalization_40[0][0]
__________________________________________________________________________________________________
activation_43 (Activation) (None, 17, 17, 192) 0 batch_normalization_43[0][0]
__________________________________________________________________________________________________
activation_48 (Activation) (None, 17, 17, 192) 0 batch_normalization_48[0][0]
__________________________________________________________________________________________________
activation_49 (Activation) (None, 17, 17, 192) 0 batch_normalization_49[0][0]
__________________________________________________________________________________________________
mixed5 (Concatenate) (None, 17, 17, 768) 0 activation_40[0][0]
activation_43[0][0]
activation_48[0][0]
activation_49[0][0]
__________________________________________________________________________________________________
conv2d_54 (Conv2D) (None, 17, 17, 160) 122880 mixed5[0][0]
__________________________________________________________________________________________________
batch_normalization_54 (BatchNo (None, 17, 17, 160) 480 conv2d_54[0][0]
__________________________________________________________________________________________________
activation_54 (Activation) (None, 17, 17, 160) 0 batch_normalization_54[0][0]
__________________________________________________________________________________________________
conv2d_55 (Conv2D) (None, 17, 17, 160) 179200 activation_54[0][0]
__________________________________________________________________________________________________
batch_normalization_55 (BatchNo (None, 17, 17, 160) 480 conv2d_55[0][0]
__________________________________________________________________________________________________
activation_55 (Activation) (None, 17, 17, 160) 0 batch_normalization_55[0][0]
__________________________________________________________________________________________________
conv2d_51 (Conv2D) (None, 17, 17, 160) 122880 mixed5[0][0]
__________________________________________________________________________________________________
conv2d_56 (Conv2D) (None, 17, 17, 160) 179200 activation_55[0][0]
__________________________________________________________________________________________________
batch_normalization_51 (BatchNo (None, 17, 17, 160) 480 conv2d_51[0][0]
__________________________________________________________________________________________________
batch_normalization_56 (BatchNo (None, 17, 17, 160) 480 conv2d_56[0][0]
__________________________________________________________________________________________________
activation_51 (Activation) (None, 17, 17, 160) 0 batch_normalization_51[0][0]
__________________________________________________________________________________________________
activation_56 (Activation) (None, 17, 17, 160) 0 batch_normalization_56[0][0]
__________________________________________________________________________________________________
conv2d_52 (Conv2D) (None, 17, 17, 160) 179200 activation_51[0][0]
__________________________________________________________________________________________________
conv2d_57 (Conv2D) (None, 17, 17, 160) 179200 activation_56[0][0]
__________________________________________________________________________________________________
batch_normalization_52 (BatchNo (None, 17, 17, 160) 480 conv2d_52[0][0]
__________________________________________________________________________________________________
batch_normalization_57 (BatchNo (None, 17, 17, 160) 480 conv2d_57[0][0]
__________________________________________________________________________________________________
activation_52 (Activation) (None, 17, 17, 160) 0 batch_normalization_52[0][0]
__________________________________________________________________________________________________
activation_57 (Activation) (None, 17, 17, 160) 0 batch_normalization_57[0][0]
__________________________________________________________________________________________________
average_pooling2d_5 (AveragePoo (None, 17, 17, 768) 0 mixed5[0][0]
__________________________________________________________________________________________________
conv2d_50 (Conv2D) (None, 17, 17, 192) 147456 mixed5[0][0]
__________________________________________________________________________________________________
conv2d_53 (Conv2D) (None, 17, 17, 192) 215040 activation_52[0][0]
__________________________________________________________________________________________________
conv2d_58 (Conv2D) (None, 17, 17, 192) 215040 activation_57[0][0]
__________________________________________________________________________________________________
conv2d_59 (Conv2D) (None, 17, 17, 192) 147456 average_pooling2d_5[0][0]
__________________________________________________________________________________________________
batch_normalization_50 (BatchNo (None, 17, 17, 192) 576 conv2d_50[0][0]
__________________________________________________________________________________________________
batch_normalization_53 (BatchNo (None, 17, 17, 192) 576 conv2d_53[0][0]
__________________________________________________________________________________________________
batch_normalization_58 (BatchNo (None, 17, 17, 192) 576 conv2d_58[0][0]
__________________________________________________________________________________________________
batch_normalization_59 (BatchNo (None, 17, 17, 192) 576 conv2d_59[0][0]
__________________________________________________________________________________________________
activation_50 (Activation) (None, 17, 17, 192) 0 batch_normalization_50[0][0]
__________________________________________________________________________________________________
activation_53 (Activation) (None, 17, 17, 192) 0 batch_normalization_53[0][0]
__________________________________________________________________________________________________
activation_58 (Activation) (None, 17, 17, 192) 0 batch_normalization_58[0][0]
__________________________________________________________________________________________________
activation_59 (Activation) (None, 17, 17, 192) 0 batch_normalization_59[0][0]
__________________________________________________________________________________________________
mixed6 (Concatenate) (None, 17, 17, 768) 0 activation_50[0][0]
activation_53[0][0]
activation_58[0][0]
activation_59[0][0]
__________________________________________________________________________________________________
conv2d_64 (Conv2D) (None, 17, 17, 192) 147456 mixed6[0][0]
__________________________________________________________________________________________________
batch_normalization_64 (BatchNo (None, 17, 17, 192) 576 conv2d_64[0][0]
__________________________________________________________________________________________________
activation_64 (Activation) (None, 17, 17, 192) 0 batch_normalization_64[0][0]
__________________________________________________________________________________________________
conv2d_65 (Conv2D) (None, 17, 17, 192) 258048 activation_64[0][0]
__________________________________________________________________________________________________
batch_normalization_65 (BatchNo (None, 17, 17, 192) 576 conv2d_65[0][0]
__________________________________________________________________________________________________
activation_65 (Activation) (None, 17, 17, 192) 0 batch_normalization_65[0][0]
__________________________________________________________________________________________________
conv2d_61 (Conv2D) (None, 17, 17, 192) 147456 mixed6[0][0]
__________________________________________________________________________________________________
conv2d_66 (Conv2D) (None, 17, 17, 192) 258048 activation_65[0][0]
__________________________________________________________________________________________________
batch_normalization_61 (BatchNo (None, 17, 17, 192) 576 conv2d_61[0][0]
__________________________________________________________________________________________________
batch_normalization_66 (BatchNo (None, 17, 17, 192) 576 conv2d_66[0][0]
__________________________________________________________________________________________________
activation_61 (Activation) (None, 17, 17, 192) 0 batch_normalization_61[0][0]
__________________________________________________________________________________________________
activation_66 (Activation) (None, 17, 17, 192) 0 batch_normalization_66[0][0]
__________________________________________________________________________________________________
conv2d_62 (Conv2D) (None, 17, 17, 192) 258048 activation_61[0][0]
__________________________________________________________________________________________________
conv2d_67 (Conv2D) (None, 17, 17, 192) 258048 activation_66[0][0]
__________________________________________________________________________________________________
batch_normalization_62 (BatchNo (None, 17, 17, 192) 576 conv2d_62[0][0]
__________________________________________________________________________________________________
batch_normalization_67 (BatchNo (None, 17, 17, 192) 576 conv2d_67[0][0]
__________________________________________________________________________________________________
activation_62 (Activation) (None, 17, 17, 192) 0 batch_normalization_62[0][0]
__________________________________________________________________________________________________
activation_67 (Activation) (None, 17, 17, 192) 0 batch_normalization_67[0][0]
__________________________________________________________________________________________________
average_pooling2d_6 (AveragePoo (None, 17, 17, 768) 0 mixed6[0][0]
__________________________________________________________________________________________________
conv2d_60 (Conv2D) (None, 17, 17, 192) 147456 mixed6[0][0]
__________________________________________________________________________________________________
conv2d_63 (Conv2D) (None, 17, 17, 192) 258048 activation_62[0][0]
__________________________________________________________________________________________________
conv2d_68 (Conv2D) (None, 17, 17, 192) 258048 activation_67[0][0]
__________________________________________________________________________________________________
conv2d_69 (Conv2D) (None, 17, 17, 192) 147456 average_pooling2d_6[0][0]
__________________________________________________________________________________________________
batch_normalization_60 (BatchNo (None, 17, 17, 192) 576 conv2d_60[0][0]
__________________________________________________________________________________________________
batch_normalization_63 (BatchNo (None, 17, 17, 192) 576 conv2d_63[0][0]
__________________________________________________________________________________________________
batch_normalization_68 (BatchNo (None, 17, 17, 192) 576 conv2d_68[0][0]
__________________________________________________________________________________________________
batch_normalization_69 (BatchNo (None, 17, 17, 192) 576 conv2d_69[0][0]
__________________________________________________________________________________________________
activation_60 (Activation) (None, 17, 17, 192) 0 batch_normalization_60[0][0]
__________________________________________________________________________________________________
activation_63 (Activation) (None, 17, 17, 192) 0 batch_normalization_63[0][0]
__________________________________________________________________________________________________
activation_68 (Activation) (None, 17, 17, 192) 0 batch_normalization_68[0][0]
__________________________________________________________________________________________________
activation_69 (Activation) (None, 17, 17, 192) 0 batch_normalization_69[0][0]
__________________________________________________________________________________________________
mixed7 (Concatenate) (None, 17, 17, 768) 0 activation_60[0][0]
activation_63[0][0]
activation_68[0][0]
activation_69[0][0]
__________________________________________________________________________________________________
conv2d_72 (Conv2D) (None, 17, 17, 192) 147456 mixed7[0][0]
__________________________________________________________________________________________________
batch_normalization_72 (BatchNo (None, 17, 17, 192) 576 conv2d_72[0][0]
__________________________________________________________________________________________________
activation_72 (Activation) (None, 17, 17, 192) 0 batch_normalization_72[0][0]
__________________________________________________________________________________________________
conv2d_73 (Conv2D) (None, 17, 17, 192) 258048 activation_72[0][0]
__________________________________________________________________________________________________
batch_normalization_73 (BatchNo (None, 17, 17, 192) 576 conv2d_73[0][0]
__________________________________________________________________________________________________
activation_73 (Activation) (None, 17, 17, 192) 0 batch_normalization_73[0][0]
__________________________________________________________________________________________________
conv2d_70 (Conv2D) (None, 17, 17, 192) 147456 mixed7[0][0]
__________________________________________________________________________________________________
conv2d_74 (Conv2D) (None, 17, 17, 192) 258048 activation_73[0][0]
__________________________________________________________________________________________________
batch_normalization_70 (BatchNo (None, 17, 17, 192) 576 conv2d_70[0][0]
__________________________________________________________________________________________________
batch_normalization_74 (BatchNo (None, 17, 17, 192) 576 conv2d_74[0][0]
__________________________________________________________________________________________________
activation_70 (Activation) (None, 17, 17, 192) 0 batch_normalization_70[0][0]
__________________________________________________________________________________________________
activation_74 (Activation) (None, 17, 17, 192) 0 batch_normalization_74[0][0]
__________________________________________________________________________________________________
conv2d_71 (Conv2D) (None, 8, 8, 320) 552960 activation_70[0][0]
__________________________________________________________________________________________________
conv2d_75 (Conv2D) (None, 8, 8, 192) 331776 activation_74[0][0]
__________________________________________________________________________________________________
batch_normalization_71 (BatchNo (None, 8, 8, 320) 960 conv2d_71[0][0]
__________________________________________________________________________________________________
batch_normalization_75 (BatchNo (None, 8, 8, 192) 576 conv2d_75[0][0]
__________________________________________________________________________________________________
activation_71 (Activation) (None, 8, 8, 320) 0 batch_normalization_71[0][0]
__________________________________________________________________________________________________
activation_75 (Activation) (None, 8, 8, 192) 0 batch_normalization_75[0][0]
__________________________________________________________________________________________________
max_pooling2d_3 (MaxPooling2D) (None, 8, 8, 768) 0 mixed7[0][0]
__________________________________________________________________________________________________
mixed8 (Concatenate) (None, 8, 8, 1280) 0 activation_71[0][0]
activation_75[0][0]
max_pooling2d_3[0][0]
__________________________________________________________________________________________________
conv2d_80 (Conv2D) (None, 8, 8, 448) 573440 mixed8[0][0]
__________________________________________________________________________________________________
batch_normalization_80 (BatchNo (None, 8, 8, 448) 1344 conv2d_80[0][0]
__________________________________________________________________________________________________
activation_80 (Activation) (None, 8, 8, 448) 0 batch_normalization_80[0][0]
__________________________________________________________________________________________________
conv2d_77 (Conv2D) (None, 8, 8, 384) 491520 mixed8[0][0]
__________________________________________________________________________________________________
conv2d_81 (Conv2D) (None, 8, 8, 384) 1548288 activation_80[0][0]
__________________________________________________________________________________________________
batch_normalization_77 (BatchNo (None, 8, 8, 384) 1152 conv2d_77[0][0]
__________________________________________________________________________________________________
batch_normalization_81 (BatchNo (None, 8, 8, 384) 1152 conv2d_81[0][0]
__________________________________________________________________________________________________
activation_77 (Activation) (None, 8, 8, 384) 0 batch_normalization_77[0][0]
__________________________________________________________________________________________________
activation_81 (Activation) (None, 8, 8, 384) 0 batch_normalization_81[0][0]
__________________________________________________________________________________________________
conv2d_78 (Conv2D) (None, 8, 8, 384) 442368 activation_77[0][0]
__________________________________________________________________________________________________
conv2d_79 (Conv2D) (None, 8, 8, 384) 442368 activation_77[0][0]
__________________________________________________________________________________________________
conv2d_82 (Conv2D) (None, 8, 8, 384) 442368 activation_81[0][0]
__________________________________________________________________________________________________
conv2d_83 (Conv2D) (None, 8, 8, 384) 442368 activation_81[0][0]
__________________________________________________________________________________________________
average_pooling2d_7 (AveragePoo (None, 8, 8, 1280) 0 mixed8[0][0]
__________________________________________________________________________________________________
conv2d_76 (Conv2D) (None, 8, 8, 320) 409600 mixed8[0][0]
__________________________________________________________________________________________________
batch_normalization_78 (BatchNo (None, 8, 8, 384) 1152 conv2d_78[0][0]
__________________________________________________________________________________________________
batch_normalization_79 (BatchNo (None, 8, 8, 384) 1152 conv2d_79[0][0]
__________________________________________________________________________________________________
batch_normalization_82 (BatchNo (None, 8, 8, 384) 1152 conv2d_82[0][0]
__________________________________________________________________________________________________
batch_normalization_83 (BatchNo (None, 8, 8, 384) 1152 conv2d_83[0][0]
__________________________________________________________________________________________________
conv2d_84 (Conv2D) (None, 8, 8, 192) 245760 average_pooling2d_7[0][0]
__________________________________________________________________________________________________
batch_normalization_76 (BatchNo (None, 8, 8, 320) 960 conv2d_76[0][0]
__________________________________________________________________________________________________
activation_78 (Activation) (None, 8, 8, 384) 0 batch_normalization_78[0][0]
__________________________________________________________________________________________________
activation_79 (Activation) (None, 8, 8, 384) 0 batch_normalization_79[0][0]
__________________________________________________________________________________________________
activation_82 (Activation) (None, 8, 8, 384) 0 batch_normalization_82[0][0]
__________________________________________________________________________________________________
activation_83 (Activation) (None, 8, 8, 384) 0 batch_normalization_83[0][0]
__________________________________________________________________________________________________
batch_normalization_84 (BatchNo (None, 8, 8, 192) 576 conv2d_84[0][0]
__________________________________________________________________________________________________
activation_76 (Activation) (None, 8, 8, 320) 0 batch_normalization_76[0][0]
__________________________________________________________________________________________________
mixed9_0 (Concatenate) (None, 8, 8, 768) 0 activation_78[0][0]
activation_79[0][0]
__________________________________________________________________________________________________
concatenate (Concatenate) (None, 8, 8, 768) 0 activation_82[0][0]
activation_83[0][0]
__________________________________________________________________________________________________
activation_84 (Activation) (None, 8, 8, 192) 0 batch_normalization_84[0][0]
__________________________________________________________________________________________________
mixed9 (Concatenate) (None, 8, 8, 2048) 0 activation_76[0][0]
mixed9_0[0][0]
concatenate[0][0]
activation_84[0][0]
__________________________________________________________________________________________________
conv2d_89 (Conv2D) (None, 8, 8, 448) 917504 mixed9[0][0]
__________________________________________________________________________________________________
batch_normalization_89 (BatchNo (None, 8, 8, 448) 1344 conv2d_89[0][0]
__________________________________________________________________________________________________
activation_89 (Activation) (None, 8, 8, 448) 0 batch_normalization_89[0][0]
__________________________________________________________________________________________________
conv2d_86 (Conv2D) (None, 8, 8, 384) 786432 mixed9[0][0]
__________________________________________________________________________________________________
conv2d_90 (Conv2D) (None, 8, 8, 384) 1548288 activation_89[0][0]
__________________________________________________________________________________________________
batch_normalization_86 (BatchNo (None, 8, 8, 384) 1152 conv2d_86[0][0]
__________________________________________________________________________________________________
batch_normalization_90 (BatchNo (None, 8, 8, 384) 1152 conv2d_90[0][0]
__________________________________________________________________________________________________
activation_86 (Activation) (None, 8, 8, 384) 0 batch_normalization_86[0][0]
__________________________________________________________________________________________________
activation_90 (Activation) (None, 8, 8, 384) 0 batch_normalization_90[0][0]
__________________________________________________________________________________________________
conv2d_87 (Conv2D) (None, 8, 8, 384) 442368 activation_86[0][0]
__________________________________________________________________________________________________
conv2d_88 (Conv2D) (None, 8, 8, 384) 442368 activation_86[0][0]
__________________________________________________________________________________________________
conv2d_91 (Conv2D) (None, 8, 8, 384) 442368 activation_90[0][0]
__________________________________________________________________________________________________
conv2d_92 (Conv2D) (None, 8, 8, 384) 442368 activation_90[0][0]
__________________________________________________________________________________________________
average_pooling2d_8 (AveragePoo (None, 8, 8, 2048) 0 mixed9[0][0]
__________________________________________________________________________________________________
conv2d_85 (Conv2D) (None, 8, 8, 320) 655360 mixed9[0][0]
__________________________________________________________________________________________________
batch_normalization_87 (BatchNo (None, 8, 8, 384) 1152 conv2d_87[0][0]
__________________________________________________________________________________________________
batch_normalization_88 (BatchNo (None, 8, 8, 384) 1152 conv2d_88[0][0]
__________________________________________________________________________________________________
batch_normalization_91 (BatchNo (None, 8, 8, 384) 1152 conv2d_91[0][0]
__________________________________________________________________________________________________
batch_normalization_92 (BatchNo (None, 8, 8, 384) 1152 conv2d_92[0][0]
__________________________________________________________________________________________________
conv2d_93 (Conv2D) (None, 8, 8, 192) 393216 average_pooling2d_8[0][0]
__________________________________________________________________________________________________
batch_normalization_85 (BatchNo (None, 8, 8, 320) 960 conv2d_85[0][0]
__________________________________________________________________________________________________
activation_87 (Activation) (None, 8, 8, 384) 0 batch_normalization_87[0][0]
__________________________________________________________________________________________________
activation_88 (Activation) (None, 8, 8, 384) 0 batch_normalization_88[0][0]
__________________________________________________________________________________________________
activation_91 (Activation) (None, 8, 8, 384) 0 batch_normalization_91[0][0]
__________________________________________________________________________________________________
activation_92 (Activation) (None, 8, 8, 384) 0 batch_normalization_92[0][0]
__________________________________________________________________________________________________
batch_normalization_93 (BatchNo (None, 8, 8, 192) 576 conv2d_93[0][0]
__________________________________________________________________________________________________
activation_85 (Activation) (None, 8, 8, 320) 0 batch_normalization_85[0][0]
__________________________________________________________________________________________________
mixed9_1 (Concatenate) (None, 8, 8, 768) 0 activation_87[0][0]
activation_88[0][0]
__________________________________________________________________________________________________
concatenate_1 (Concatenate) (None, 8, 8, 768) 0 activation_91[0][0]
activation_92[0][0]
__________________________________________________________________________________________________
activation_93 (Activation) (None, 8, 8, 192) 0 batch_normalization_93[0][0]
__________________________________________________________________________________________________
mixed10 (Concatenate) (None, 8, 8, 2048) 0 activation_85[0][0]
mixed9_1[0][0]
concatenate_1[0][0]
activation_93[0][0]
__________________________________________________________________________________________________
avg_pool (GlobalAveragePooling2 (None, 2048) 0 mixed10[0][0]
__________________________________________________________________________________________________
predictions (Dense) (None, 1000) 2049000 avg_pool[0][0]
==================================================================================================
Total params: 23,851,784
Trainable params: 23,817,352
Non-trainable params: 34,432
_________________________________________________________
References
Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jonathon Shlens, & Zbigniew Wojna (2016). Rethinking the Inception Architecture for Computer Vision computer vision and pattern recognition.
