resnet代码分析-pytorch和keras实现

keras

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue Apr  6 15:09:56 2021

@author: Lihuanyu
"""

from __future__ import print_function

from keras import layers
from keras.layers import (Activation, AveragePooling2D, BatchNormalization,
                          Conv2D, Dense, Flatten, Input, 
                          )
from keras.models import Model


def identity_block(input_tensor, kernel_size, filters):   #输入输出相同
    filters1, filters2, filters3 = filters


    x = Conv2D(filters1, (1, 1), use_bias=True)(input_tensor)  #进行1×1的卷积
    x = BatchNormalization()(x)   #BN
    x = Activation('relu')(x)   #激活

    x = Conv2D(filters2, kernel_size,padding='same', use_bias=True)(x)    #进行3×3的卷积
    x = BatchNormalization()(x)  #BN
    x = Activation('relu')(x)

    x = Conv2D(filters3, (1, 1), use_bias=True)(x)  #进行1×1的卷积
    x = BatchNormalization()(x)   #BN

    x = layers.add([x, input_tensor])  
    x = Activation('relu')(x)
    return x

# 输入张量，卷积核，滤波器[filters1,filters2,filters3]，stride
def conv_block(input_tensor, kernel_size, filters, strides=(2, 2)):
    filters1, filters2, filters3 = filters

    x = Conv2D(filters1, (1, 1), use_bias=True)(input_tensor)
    x = BatchNormalization()(x)
    x = Activation('relu')(x)

    x = Conv2D(filters2, kernel_size, strides=strides, padding='same', use_bias=True)(x)
    x = BatchNormalization()(x)
    x = Activation('relu')(x)


    x = Conv2D(filters3, (1, 1), use_bias=True)(x)
    x = BatchNormalization()(x)


    shortcut = Conv2D(filters3, (1, 1), strides=strides, use_bias=True)(input_tensor)
    shortcut = BatchNormalization()(shortcut)

    x = layers.add([x, shortcut])
    x = Activation('relu')(x)
    return x


def ResNet50(input_shape=[32,32,3],classes=10):
    img_input = Input(shape=input_shape)

    # 32,32,3 -> 32,32,64   进行3×3的卷积
    x = Conv2D(64, (3, 3), strides=(1, 1), padding="same", use_bias=True)(img_input)
    x = BatchNormalization()(x)
    x = Activation('relu')(x)
    

    # 32,32,64 -> 32,32,256
    x = conv_block(x, 3, [64, 64, 256], strides=(1, 1))
    x = identity_block(x, 3, [64, 64, 256])
    x = identity_block(x, 3, [64, 64, 256])
    

    # 32,32,256 -> 16,16,512
    x = conv_block(x, 3, [128, 128, 512])
    x = identity_block(x, 3, [128, 128, 512])
    x = identity_block(x, 3, [128, 128, 512])
    x = identity_block(x, 3, [128, 128, 512])

    # 16,16,512 -> 8,8,1024
    x = conv_block(x, 3, [256, 256, 1024])
    x = identity_block(x, 3, [256, 256, 1024])
    x = identity_block(x, 3, [256, 256, 1024])
    x = identity_block(x, 3, [256, 256, 1024])
    x = identity_block(x, 3, [256, 256, 1024])
    x = identity_block(x, 3, [256, 256, 1024])

    # 8,8,1024 -> 4,4,2048
    x = conv_block(x, 3, [512, 512, 2048])
    x = identity_block(x, 3, [512, 512, 2048])
    x = identity_block(x, 3, [512, 512, 2048])

    # 4,4,2048 -> 1,1,2048
    x = AveragePooling2D((4, 4), name='avg_pool')(x)

    # 1,1,2048 -> 2048 -> num_classes
    x = Flatten()(x)
    x = Dense(classes, activation='softmax', name='fc1000')(x)

    model = Model(img_input, x, name='resnet50')
    return model


if __name__ == '__main__':
    model = ResNet50()
    model.summary()

pytorch

'''ResNet in PyTorch.

For Pre-activation ResNet, see 'preact_resnet.py'.

Reference:
[1] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
    Deep Residual Learning for Image Recognition. arXiv:1512.03385
'''
import torch
import torch.nn as nn
import torch.nn.functional as F


class BasicBlock(nn.Module):
    expansion = 1

    def __init__(self, in_planes, planes, stride=1):
        super(BasicBlock, self).__init__()
        self.conv1 = nn.Conv2d(
            in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3,
                               stride=1, padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(planes)

        self.shortcut = nn.Sequential()
        if stride != 1 or in_planes != self.expansion*planes:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_planes, self.expansion*planes,
                          kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(self.expansion*planes)
            )

    def forward(self, x):
        out = F.relu(self.bn1(self.conv1(x)))
        out = self.bn2(self.conv2(out))
        out += self.shortcut(x)
        out = F.relu(out)
        return out


class Bottleneck(nn.Module):
    expansion = 4

    def __init__(self, in_planes, planes, stride=1):
        super(Bottleneck, self).__init__()
        self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3,
                               stride=stride, padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(planes)
        self.conv3 = nn.Conv2d(planes, self.expansion *
                               planes, kernel_size=1, bias=False)
        self.bn3 = nn.BatchNorm2d(self.expansion*planes)

        self.shortcut = nn.Sequential()
        if stride != 1 or in_planes != self.expansion*planes:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_planes, self.expansion*planes,
                          kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(self.expansion*planes)
            )

    def forward(self, x):
        out = F.relu(self.bn1(self.conv1(x)))
        out = F.relu(self.bn2(self.conv2(out)))
        out = self.bn3(self.conv3(out))
        out += self.shortcut(x)
        out = F.relu(out)
        return out


class ResNet(nn.Module):
    def __init__(self, block, num_blocks, num_classes=10):
        super(ResNet, self).__init__()
        self.in_planes = 64

        self.conv1 = nn.Conv2d(3, 64, kernel_size=3,
                               stride=1, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(64)
        self.layer1 = self._make_layer(block, 64, num_blocks[0], stride=1)
        self.layer2 = self._make_layer(block, 128, num_blocks[1], stride=2)
        self.layer3 = self._make_layer(block, 256, num_blocks[2], stride=2)
        self.layer4 = self._make_layer(block, 512, num_blocks[3], stride=2)
        self.linear = nn.Linear(512*block.expansion, num_classes)

    def _make_layer(self, block, planes, num_blocks, stride):
        strides = [stride] + [1]*(num_blocks-1)
        layers = []
        for stride in strides:
            layers.append(block(self.in_planes, planes, stride))
            self.in_planes = planes * block.expansion
        return nn.Sequential(*layers)

    def forward(self, x):
        out = F.relu(self.bn1(self.conv1(x)))
        out = self.layer1(out)
        out = self.layer2(out)
        out = self.layer3(out)
        out = self.layer4(out)
        out = F.avg_pool2d(out, 4)
        out = out.view(out.size(0), -1)
        out = self.linear(out)
        return out


def ResNet18():
    return ResNet(BasicBlock, [2, 2, 2, 2])


def ResNet34():
    return ResNet(BasicBlock, [3, 4, 6, 3])


def ResNet50():
    return ResNet(Bottleneck, [3, 4, 6, 3])


def ResNet101():
    return ResNet(Bottleneck, [3, 4, 23, 3])


def ResNet152():
    return ResNet(Bottleneck, [3, 8, 36, 3])


def test():
    net = ResNet18()
    y = net(torch.randn(1, 3, 32, 32))
    print(y.size())

# test()

问题：

pytorch 57、 87-93

stride != 1

 def _make_layer(self, block, planes, num_blocks, stride):
        strides = [stride] + [1]*(num_blocks-1)
        layers = []
        for stride in strides:
            layers.append(block(self.in_planes, planes, stride))
            self.in_planes = planes * block.expansion
        return nn.Sequential(*layers)