Embedded Files
1) Question:
4.1 LetNet5 129
4.1.1 Model Structure 129
4.1.2 Model Structure 129
4.1.3 Model characteristics 131
4.2 AlexNet 131
4.2.1 Model structure 131
4.2.2 Model Interpretation 131
4.2.3 Model characteristics 135
4.3 Visualization ZFNet-Deconvolution 135
4.3.1 Basic ideas and processes 135
4.3.2 Convolution and Deconvolution 136
4.3.3 Convolution Visualization 137
4.3.4 Comparison of ZFNe and AlexNet 139
4.4 VGG 140
4.1.1 Model Structure 140
4.1.2 Model Features 140
4.5 Network in Network 141
4.5.1 Model Structure 141
4.5.2 Model Innovation Points 141
4.6 GoogleNet 143
4.6.1 Model Structure 143
4.6.2 Inception Structure 145
4.6.3 Model hierarchy 146
4.7 Inception Series 148
4.7.1 Inception v1 148
4.7.2 Inception v2 150
4.7.3 Inception v3 153
4.7.4 Inception V4 155
4.7.5 Inception-ResNet-v2 157
4.8 ResNet and its variants 158
4.8.1 Reviewing ResNet 159
4.8.2 residual block 160
4.8.3 ResNet Architecture 162
4.8.4 Variants of residual blocks 162
4.8.5 ResNeXt 162
4.8.6 Densely Connected CNN 164
4.8.7 ResNet as a combination of small networks 165
4.8.8 Features of Paths in ResNet 166
4.9 Why are the current CNN models adjusted on GoogleNet, VGGNet or AlexNet? 167
2) Answer:
P
References:
Page updated
Google Sites
Report abuse