HarukiYqM

## resnet-152_keras.py
# -*- coding: utf-8 -*-

import cv2
import numpy as np
import copy

from keras.layers import Input, Dense, Conv2D, MaxPooling2D, AveragePooling2D, ZeroPadding2D, Flatten, Activation, add
from keras.optimizers import SGD
from keras.layers.normalization import BatchNormalization
from keras.models import Model

## readme.md

      
        
          
            
              
              2 files
            
          
          
            
              
              0 forks
            
          
            
              
                
                0 comments
              
            
          
            
              
              0 stars
            
          
        
        
          
              
          
          
            
                HarukiYqM
                / readme.md
            
            
              Created
              April 15, 2018 04:16
                — forked from flyyufelix/readme.md
            
              
                Resnet-152 pre-trained model in Keras
              
          
        
      
        
  
      
    ResNet-152 in Keras

This is an Keras implementation of ResNet-152 with ImageNet pre-trained weights. I converted the weights from Caffe provided by the authors of the paper. The implementation supports both Theano and TensorFlow backends. Just in case you are curious about how the conversion is done, you can visit my blog post for more details.
ResNet Paper:
Deep Residual Learning for Image Recognition.
Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
arXiv:1512.03385
	# -- coding: utf-8 --

	import cv2
	import numpy as np
	import copy

	from keras.layers import Input, Dense, Conv2D, MaxPooling2D, AveragePooling2D, ZeroPadding2D, Flatten, Activation, add
	from keras.optimizers import SGD
	from keras.layers.normalization import BatchNormalization
	from keras.models import Model