Please choose an image file to be uploaded


AlexNet

trained on ILSVRC 2012


                

Fine-tuning CaffeNet for Style Recognition

trained on “Flickr Style” Data


                

GoogLeNet

official model trained on ILSVRC 2012


                

                

CaffeNet

AlexNet trained on ILSVRC 2012, with a minor variation from the version as described in ImageNet classification with deep convolutional neural networks by Krizhevsky et al. in NIPS 2012.


                

ResNet

ResNet-152 described in the paper "Deep Residual Learning for Image Recognition".
Used in ILSVRC and COCO 2015 competitions, which won the 1st places in: ImageNet classification, ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

            

各模型对比及在ILSVRC的历年Top-5错误率

Model Name AlexNet VGG GoogLeNet ResNet
Year 2012 2014 2014 2015
Layers 8 19 22 152
Top-5 Error 16.4% 7.3% 6.7% 3.57%
Data Augmentation + + + +
Inception(NIN) +
Conv Layers 5 16 21 151
Conv Kernel Sizes 11,5,3 3 7,1,3,5 7,1,3,5
Fully-Connected Layers 3 3 1 1
Fully-Connected Layer Sizes 4096,4096,1000 4096,4096,1000 1000 1000
Dropout + + + +
Local Response Normalization + +
Batch Normalization +