1. 概述

1.1 论文下载

EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

原址下载
百度云下载，提取码

1.2 论文解读

参考：https://blog.csdn.net/qq_37541097/article/details/114434046

2.性能对比

3. 网络backbone

Model	input_size	width_coefficient	depth_coefficient	drop_connect_rate	dropout_rate
EfficientNetB0	224x224	1.0	1.0	0.2	0.2
EfficientNetB1	240x240	1.0	1.1	0.2	0.2
EfficientNetB2	260x260	1.1	1.2	0.2	0.3
EfficientNetB3	300x300	1.2	1.4	0.2	0.3
EfficientNetB4	380x380	1.4	1.8	0.2	0.4
EfficientNetB5	456x456	1.6	2.2	0.2	0.4
EfficientNetB6	528x528	1.8	2.6	0.2	0.5
EfficientNetB7	600x600	2.0	3.1	0.2	0.5

3.1 MBcov模块

3.2 SE模块

4. 代码实现

import math
import copy
from functools import partial
from collections import OrderedDict
from typing import Optional, Callableimport torch
import torch.nn as nn
from torch import Tensor
from torch.nn import functional as F"""
该方法就是将传入的channel的个数调整到8的整数倍，使得对硬件友好
"""
def _make_divisible(ch, divisor=8, min_ch=None):"""This function is taken from the original tf repo.It ensures that all layers have a channel number that is divisible by 8It can be seen here:https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py"""if min_ch is None:min_ch = divisornew_ch = max(min_ch, int(ch + divisor / 2) // divisor * divisor)# Make sure that round down does not go down by more than 10%.if new_ch < 0.9 * ch:new_ch += divisorreturn new_ch""""""
def drop_path(x, drop_prob: float = 0., training: bool = False):"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks)."Deep Networks with Stochastic Depth", https://arxiv.org/pdf/1603.09382.pdfThis function is taken from the rwightman.It can be seen here:https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/drop.py#L140"""if drop_prob == 0. or not training:return xkeep_prob = 1 - drop_probshape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNetsrandom_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device)random_tensor.floor_()  # binarizeoutput = x.div(keep_prob) * random_tensorreturn output""""""
class DropPath(nn.Module):"""Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks)."Deep Networks with Stochastic Depth", https://arxiv.org/pdf/1603.09382.pdf"""def __init__(self, drop_prob=None):super(DropPath, self).__init__()self.drop_prob = drop_probdef forward(self, x):return drop_path(x, self.drop_prob, self.training)"""
定义BN和激活函数
"""
class ConvBNActivation(nn.Sequential):def __init__(self,in_planes: int,out_planes: int,kernel_size: int = 3,stride: int = 1,groups: int = 1,norm_layer: Optional[Callable[..., nn.Module]] = None, #BN结构，默认为Noneactivation_layer: Optional[Callable[..., nn.Module]] = None): #ac结构，默认为Nonepadding = (kernel_size - 1) // 2if norm_layer is None:"""不指定norm的话，默认BN结构"""norm_layer = nn.BatchNorm2dif activation_layer is None:"""如果不指定激活函数，则默认SiLU激活函数"""activation_layer = nn.SiLU  # alias Swish  (torch>=1.7)激活函数使用的是SiLU，在版本1.70以上才有，建议1.7.1版本#定义层结构[in_channels,out_channels,kernel_size,stride,padding,groups,bias=False],-->BN-->acsuper(ConvBNActivation, self).__init__(nn.Conv2d(in_channels=in_planes,out_channels=out_planes,kernel_size=kernel_size,stride=stride,padding=padding,groups=groups,bias=False),norm_layer(out_planes),activation_layer())"""
定义SE模块
"""
class SqueezeExcitation(nn.Module):def __init__(self,input_c: int,   # block input channelexpand_c: int,  # block expand channelsqueeze_factor: int = 4):super(SqueezeExcitation, self).__init__()squeeze_c = input_c // squeeze_factorself.fc1 = nn.Conv2d(expand_c, squeeze_c, 1)self.ac1 = nn.SiLU()  # alias Swishself.fc2 = nn.Conv2d(squeeze_c, expand_c, 1)self.ac2 = nn.Sigmoid()def forward(self, x: Tensor) -> Tensor:scale = F.adaptive_avg_pool2d(x, output_size=(1, 1))scale = self.fc1(scale)scale = self.ac1(scale)scale = self.fc2(scale)scale = self.ac2(scale)return scale * x""""""
class InvertedResidualConfig:# kernel_size, in_channel, out_channel, exp_ratio, strides, use_SE, drop_connect_ratedef __init__(self,kernel: int,          # 3 or 5input_c: int,out_c: int,expanded_ratio: int,  # 1 or 6stride: int,          # 1 or 2use_se: bool,         # Truedrop_rate: float,index: str,           # 1a, 2a, 2b, ...width_coefficient: float):self.input_c = self.adjust_channels(input_c, width_coefficient)self.kernel = kernelself.expanded_c = self.input_c * expanded_ratioself.out_c = self.adjust_channels(out_c, width_coefficient)self.use_se = use_seself.stride = strideself.drop_rate = drop_rateself.index = index@staticmethoddef adjust_channels(channels: int, width_coefficient: float):return _make_divisible(channels * width_coefficient, 8)""""""
class InvertedResidual(nn.Module):def __init__(self,cnf: InvertedResidualConfig,norm_layer: Callable[..., nn.Module]):super(InvertedResidual, self).__init__()if cnf.stride not in [1, 2]:raise ValueError("illegal stride value.")self.use_res_connect = (cnf.stride == 1 and cnf.input_c == cnf.out_c)layers = OrderedDict()activation_layer = nn.SiLU  # alias Swish# expandif cnf.expanded_c != cnf.input_c:layers.update({"expand_conv": ConvBNActivation(cnf.input_c,cnf.expanded_c,kernel_size=1,norm_layer=norm_layer,activation_layer=activation_layer)})# depthwiselayers.update({"dwconv": ConvBNActivation(cnf.expanded_c,cnf.expanded_c,kernel_size=cnf.kernel,stride=cnf.stride,groups=cnf.expanded_c,norm_layer=norm_layer,activation_layer=activation_layer)})if cnf.use_se:layers.update({"se": SqueezeExcitation(cnf.input_c,cnf.expanded_c)})# projectlayers.update({"project_conv": ConvBNActivation(cnf.expanded_c,cnf.out_c,kernel_size=1,norm_layer=norm_layer,activation_layer=nn.Identity)})self.block = nn.Sequential(layers)self.out_channels = cnf.out_cself.is_strided = cnf.stride > 1# 只有在使用shortcut连接时才使用dropout层if self.use_res_connect and cnf.drop_rate > 0:self.dropout = DropPath(cnf.drop_rate)else:self.dropout = nn.Identity()def forward(self, x: Tensor) -> Tensor:result = self.block(x)result = self.dropout(result)if self.use_res_connect:result += xreturn result""""""
class EfficientNet(nn.Module):def __init__(self,width_coefficient: float,depth_coefficient: float,num_classes: int = 1000,dropout_rate: float = 0.2,drop_connect_rate: float = 0.2,block: Optional[Callable[..., nn.Module]] = None,norm_layer: Optional[Callable[..., nn.Module]] = None):super(EfficientNet, self).__init__()# kernel_size, in_channel, out_channel, exp_ratio, strides, use_SE, drop_connect_rate, repeatsdefault_cnf = [[3, 32, 16, 1, 1, True, drop_connect_rate, 1],[3, 16, 24, 6, 2, True, drop_connect_rate, 2],[5, 24, 40, 6, 2, True, drop_connect_rate, 2],[3, 40, 80, 6, 2, True, drop_connect_rate, 3],[5, 80, 112, 6, 1, True, drop_connect_rate, 3],[5, 112, 192, 6, 2, True, drop_connect_rate, 4],[3, 192, 320, 6, 1, True, drop_connect_rate, 1]]def round_repeats(repeats):"""Round number of repeats based on depth multiplier."""return int(math.ceil(depth_coefficient * repeats))if block is None:block = InvertedResidualif norm_layer is None:norm_layer = partial(nn.BatchNorm2d, eps=1e-3, momentum=0.1)adjust_channels = partial(InvertedResidualConfig.adjust_channels,width_coefficient=width_coefficient)# build inverted_residual_settingbneck_conf = partial(InvertedResidualConfig,width_coefficient=width_coefficient)b = 0num_blocks = float(sum(round_repeats(i[-1]) for i in default_cnf))inverted_residual_setting = []for stage, args in enumerate(default_cnf):cnf = copy.copy(args)for i in range(round_repeats(cnf.pop(-1))):if i > 0:# strides equal 1 except first cnfcnf[-3] = 1  # stridescnf[1] = cnf[2]  # input_channel equal output_channelcnf[-1] = args[-2] * b / num_blocks  # update dropout ratioindex = str(stage + 1) + chr(i + 97)  # 1a, 2a, 2b, ...inverted_residual_setting.append(bneck_conf(*cnf, index))b += 1# create layerslayers = OrderedDict()# first convlayers.update({"stem_conv": ConvBNActivation(in_planes=3,out_planes=adjust_channels(32),kernel_size=3,stride=2,norm_layer=norm_layer)})# building inverted residual blocksfor cnf in inverted_residual_setting:layers.update({cnf.index: block(cnf, norm_layer)})# build toplast_conv_input_c = inverted_residual_setting[-1].out_clast_conv_output_c = adjust_channels(1280)layers.update({"top": ConvBNActivation(in_planes=last_conv_input_c,out_planes=last_conv_output_c,kernel_size=1,norm_layer=norm_layer)})self.features = nn.Sequential(layers)self.avgpool = nn.AdaptiveAvgPool2d(1)classifier = []if dropout_rate > 0:classifier.append(nn.Dropout(p=dropout_rate, inplace=True))classifier.append(nn.Linear(last_conv_output_c, num_classes))self.classifier = nn.Sequential(*classifier)# initial weightsfor m in self.modules():if isinstance(m, nn.Conv2d):nn.init.kaiming_normal_(m.weight, mode="fan_out")if m.bias is not None:nn.init.zeros_(m.bias)elif isinstance(m, nn.BatchNorm2d):nn.init.ones_(m.weight)nn.init.zeros_(m.bias)elif isinstance(m, nn.Linear):nn.init.normal_(m.weight, 0, 0.01)nn.init.zeros_(m.bias)def _forward_impl(self, x: Tensor) -> Tensor:x = self.features(x)x = self.avgpool(x)x = torch.flatten(x, 1)x = self.classifier(x)return xdef forward(self, x: Tensor) -> Tensor:return self._forward_impl(x)""""""
def efficientnet_b0(num_classes=1000):# input image size 224x224return EfficientNet(width_coefficient=1.0,depth_coefficient=1.0,dropout_rate=0.2,num_classes=num_classes)""""""
def efficientnet_b1(num_classes=1000):# input image size 240x240return EfficientNet(width_coefficient=1.0,depth_coefficient=1.1,dropout_rate=0.2,num_classes=num_classes)
""""""
def efficientnet_b2(num_classes=1000):# input image size 260x260return EfficientNet(width_coefficient=1.1,depth_coefficient=1.2,dropout_rate=0.3,num_classes=num_classes)""""""
def efficientnet_b3(num_classes=1000):# input image size 300x300return EfficientNet(width_coefficient=1.2,depth_coefficient=1.4,dropout_rate=0.3,num_classes=num_classes)""""""
def efficientnet_b4(num_classes=1000):# input image size 380x380return EfficientNet(width_coefficient=1.4,depth_coefficient=1.8,dropout_rate=0.4,num_classes=num_classes)""""""
def efficientnet_b5(num_classes=1000):# input image size 456x456return EfficientNet(width_coefficient=1.6,depth_coefficient=2.2,dropout_rate=0.4,num_classes=num_classes)""""""
def efficientnet_b6(num_classes=1000):# input image size 528x528return EfficientNet(width_coefficient=1.8,depth_coefficient=2.6,dropout_rate=0.5,num_classes=num_classes)""""""
def efficientnet_b7(num_classes=1000):# input image size 600x600return EfficientNet(width_coefficient=2.0,depth_coefficient=3.1,dropout_rate=0.5,num_classes=num_classes)

Pytorch实现EfficientNet相关推荐

神经网络学习小记录50——Pytorch 利用efficientnet系列模型搭建yolov3目标检测平台
神经网络学习小记录50--Pytorch 利用efficientnet系列模型搭建yolov3目标检测平台学习前言什么是EfficientNet模型源码下载 EfficientNet模型的实现思 ...
【项目实战课】基于Pytorch的EfficientNet血红细胞分类竞赛实战
欢迎大家来到我们的项目实战课,本期内容是<基于Pytorch的EfficientNet血红细胞分类竞赛实战>.所谓项目课,就是以简单的原理回顾+详细的项目实战的模式,针对具体的某一个主题, ...
【小白学PyTorch】13.EfficientNet详解及PyTorch实现
<<小白学PyTorch>> 小白学PyTorch | 12 SENet详解及PyTorch实现小白学PyTorch | 11 MobileNet详解及PyTorch实现小 ...
【小白学PyTorch】扩展之Tensorflow2.0 | 21 Keras的API详解（下）池化、Normalization
<<小白学PyTorch>> 扩展之Tensorflow2.0 | 21 Keras的API详解(上)卷积.激活.初始化.正则扩展之Tensorflow2.0 | 20 TF ...
【小白学PyTorch】扩展之Tensorflow2.0 | 21 Keras的API详解（上）卷积、激活、初始化、正则...
[机器学习炼丹术]的学习笔记分享 <<小白学PyTorch>> 扩展之Tensorflow2.0 | 20 TF2的eager模式与求导扩展之Tensorflow2.0 | ...
【小白学PyTorch】扩展之Tensorflow2.0 | 20 TF2的eager模式与求导
[机器学习炼丹术]的学习笔记分享 <<小白学PyTorch>> 扩展之Tensorflow2.0 | 19 TF2模型的存储与载入扩展之Tensorflow2.0 | 18 ...
【小白学PyTorch】18.TF2构建自定义模型
[机器学习炼丹术]的学习笔记分享 <<小白学PyTorch>> 扩展之Tensorflow2.0 | 17 TFrec文件的创建与读取扩展之Tensorflow2.0 | 1 ...
【小白学PyTorch】17.TFrec文件的创建与读取
[机器学习炼丹术]的学习笔记分享 <<小白学PyTorch>> 小白学PyTorch | 16 TF2读取图片的方法小白学PyTorch | 15 TF2实现一个简单的服装分 ...
【小白学PyTorch】16.TF2读取图片的方法
<<小白学PyTorch>> 扩展之tensorflow2.0 | 15 TF2实现一个简单的服装分类任务小白学PyTorch | 14 tensorboardX可视化教程 ...
【小白学PyTorch】15.TF2实现一个简单的服装分类任务
<<小白学PyTorch>> 小白学PyTorch | 14 tensorboardX可视化教程小白学PyTorch | 13 EfficientNet详解及PyTorch实 ...

Pytorch实现EfficientNet

文章目录

1. 概述

1.1 论文下载

1.2 论文解读

2.性能对比

3. 网络backbone

3.1 MBcov模块

3.2 SE模块

4. 代码实现

Pytorch实现EfficientNet相关推荐

最新文章

热门文章