every blog every motto: You can do more than you think.

0. 前言

金字塔池化模块简单小结。

废话： 还是菜的抠脚，哎，

1. 正文

1.1 基础概念

金字塔池化模块，即用不同尺度进行池化操作。

如池化后特征图大小为1×1,（不考虑通道，因通道无变化），假若我们的输入特征图为40×40，那么，即对所有数据（1600）取平均（如下图，平均值为60）。

如池化后的特征图为2×2，那么即将特征图分成4个区域（2×2），每个区域内进行平均池化，如下图：

说明：
1. Tensorflow中，用AveragePooling2D进行池化，需要计算池化核和步幅
Pytorch中，用AdaptiveAvgPool2d，指定输出特征图大小即可
具体见后面代码
2. 上述图片中的数值为随便取得，仅展示。

金字塔池化模块，生成不同尺度的池化结果。具体结果为1×1，2×2,3×3,6×6，四种，然后将生成的特征图上采样恢复到初始特征图大小，最后在通道方向上合并（包含最初的特征图），图如下：

通俗的理解：

说明： 仅展示从特征图大小角度进行显示，未展示通道情况

1.2 代码

1.2.1 pytorch简洁版

from torch import nn
import torch
from torch.nn import functional as Fclass PPM(nn.Module):def __init__(self, in_dim, reduction_dim, bins):"""Pyramid Pooling Module 金字塔池化"""super(PPM, self).__init__()self.features = []for bin in bins:self.features.append(nn.Sequential(# 池化nn.AdaptiveAvgPool2d(bin),# 卷积，降低通道数nn.Conv2d(in_dim, reduction_dim, kernel_size=1, bias=False),nn.BatchNorm2d(reduction_dim),nn.ReLU(inplace=True)))self.features = nn.ModuleList(self.features)def forward(self, x):x_size = x.size()out = [x]for f in self.features:temp = f(x)  # 池化+卷积# print(temp.shape)# 上采样，恢复到初始大小temp = F.interpolate(temp, x_size[2:], mode='bilinear', align_corners=True)out.append(temp)return torch.cat(out, 1)  # 通道方向上合并ten = torch.rand((7, 4, 30, 30))
ppm = PPM(4, 2, [1, 2, 3, 6])
ppm(ten)

1.2.2 pytorch 老版

import torch
import torch.nn as nn
from torch.nn import functional as F
import numpy as npclass PSPModule(nn.Module):def __init__(self, in_channel, out_channel=1024, sizes=(1, 2, 3, 6)):super().__init__()self.stages = []self.stages = nn.ModuleList([self._make_stage(in_channel, size) for size in sizes])self.bottleneck = nn.Conv2d(in_channel * (len(sizes) + 1), out_channel, kernel_size=(1, 1))self.relu = nn.ReLU()def _make_stage(self, in_channel, out_size):prior = nn.AdaptiveAvgPool2d(output_size=(out_size, out_size))# print('prior shape:',prior.shape)conv = nn.Conv2d(in_channel, in_channel, kernel_size=(1, 1), bias=False)return nn.Sequential(prior, conv)def forward(self, x):h, w = x.size(2), x.size(3)box = [x]  # 收集不同的池化后特征图for layer in self.stages:tempx = layer(x)  # 自适应平均池化print('自适应平均池化 shape:', tempx.shape)x = F.upsample(input=tempx, size=(h, w), mode='bilinear')box.append(x)box = [F.upsample(input=stage(x), size=(h, w), mode='bilinear') for stage in self.stages] + [x]x = torch.cat(box, 1)  # 通道方向上合并print('通道方向合并的 shape:',x.shape)# print('shape:', priors.shape)bottle = self.bottleneck(x)return self.relu(bottle)arr = np.zeros((1, 3, 30, 30), dtype=np.float32)ten = torch.from_numpy(arr)a = PSPModule(3)
a(ten)

说明： 含有通道合并后的部分代码（卷积+relu）

1.2.3 tensorflow 2.x版

在Tensorflow中，由于没有想pytorch的自适应池化（AdaptiveAvgPool2d），所以使用平均池化（AveragePooling2D），即通过制定池化核大小和步幅,以达到同样的大小。
eg1: 输入特征图为（30,30），希望得到（2,2）的特征图（不考虑通道，因通道无变化）
那么，这个过程就是1.1中所示，pool_factor=2
池化核大小：pool_size = 30/pool_factor = 30/2 =15
步幅：strides = pool_size = 15

eg2: 希望得到（3,3）的特征图，那么pool_factor = 3
池化核大小：pool_size = 30/pool_factor = 30/3 =10
步幅：strides = pool_size = 10

eg3: 希望得到（3,3）的特征图，那么pool_factor =6
池化核大小：pool_size = 30/pool_factor = 30/6 =5
步幅：strides = pool_size = 5

注意： tensorflow中的代码都需计算这两个参数！

import osos.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import tensorflow as tf
from tensorflow.keras.layers import AveragePooling2D, Conv2D, BatchNormalization, Activation, Lambda
import tensorflow.keras.backend as K
import numpy as npdef pool_block(feats, pool_factor, out_channel):""":param feats: 输入特征图 (h,w,C):param pool_factor: 池化后大小 (h1,w1,c):param out_channel: 输出通道:return:"""print('pool factor为：', pool_factor)h = K.int_shape(feats)[1]w = K.int_shape(feats)[2]pool_size = [int(np.round(float(h) / pool_factor)), int(np.round(float(w) / pool_factor))]strides = pool_size# 不同尺度的平均池化x = AveragePooling2D(pool_size, strides=strides, padding='same')(feats)print('平均池化以后的shape:', x.shape)x = Conv2D(out_channel , (1, 1), padding='same')(x)x = BatchNormalization()(x)x = Activation('relu')(x)x = Lambda(lambda x: tf.compat.v1.image.resize_images(x, (h, w), align_corners=True))(x)print('final x shape:',x.shape)return xarr = np.zeros((1, 30, 30, 3), dtype=np.float32)
ten = tf.convert_to_tensor(arr)pool_factorts = [1, 2, 3, 6]
for p in pool_factorts:pool_block(ten, p, 60)

1.2.4 tensorflow 1.x版

import osos.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import tensorflow as tf
from tensorflow.contrib import slim
import numpy as npdef resize_image(input_images, s):""":param input_images: input:param s: 缩放比例 (2,2):return:"""h, w = input_images.get_shape().as_list()[1], input_images.get_shape().as_list()[2]h_ratio = s[0]w_ratio = s[1]h = int(h * h_ratio)w = int(w * w_ratio)images = tf.image.resize_images(input_images, size=(h, w))return imagesdef pool_block(x, pool_factor, IMAGE_ORDER='NHWC'):if IMAGE_ORDER == 'NHWC':h, w = x.get_shape().as_list()[1], x.get_shape().as_list()[2]else:h, w = x.get_shape().as_list()[2], x.get_shape().as_list()[3]# strides = [18,18],[9,9],[6,6],[3,3]pool_size = [int(np.round(float(h) / pool_factor)), int(np.round(float(w) / pool_factor))]strides = pool_size# 进行不同程度的平均池化x = slim.avg_pool2d(x, kernel_size=pool_size, stride=strides, padding='SAME')print('池化后的特征图大小:', x.shape)# 进行卷积操作x = slim.conv2d(x, 512, kernel_size=(1, 1), stride=1, padding='SAME')x = tf.cast(x, tf.float32)x = slim.batch_norm(x)x = tf.nn.relu(x)print('relu 后：', x.shape)print('-' * 100)x = resize_image(x, strides)return xarr = np.zeros((1, 30, 30, 3), dtype=np.float32)xx = tf.convert_to_tensor(arr)
pool_factors = [1, 2, 3, 6]for pf in pool_factors:p = pool_block(xx, pf)# print(p.shape)

参考文献

[1] https://blog.csdn.net/qq_24975309/article/details/108677378
[2] https://www.cnblogs.com/learningcaiji/p/14187039.html
[3] https://blog.csdn.net/qq_43258953/article/details/103300945
[4] https://zhuanlan.zhihu.com/p/75206669
[5] https://pytorch.org/docs/stable/nn.functional.html
[6] https://blog.csdn.net/weixin_44791964/article/details/108469870
[7] https://github.com/pudae/tensorflow-pspnet
[8] https://github.com/bubbliiiing/pspnet-tf2
[9] https://github.com/hszhao/semseg
[10] https://github.com/bubbliiiing/pspnet-pytorch/blob/master/nets/pspnet.py
[11] https://blog.csdn.net/qq_41375318/article/details/110071599
[12] https://www.cnblogs.com/wzyuan/p/10224793.html
[13] https://www.cnblogs.com/wzyuan/p/10224793.html
[14] https://github.com/Lextal/pspnet-pytorch/blob/master/pspnet.py

PSP模块Tensorflow/Pytorch实现小结相关推荐

【前沿】何恺明大神ICCV2017最佳论文Mask R-CNN的Keras/TensorFlow/Pytorch 代码实现
我们提出了一个概念上简单.灵活和通用的用于目标实例分割(object instance segmentation)的框架.我们的方法能够有效地检测图像中的目标,同时还能为每个实例生成一个高质量的分割掩 ...
scikit-learn, tensorflow, pytorch真的只需要查下API，不需要学吗?
链接:https://www.zhihu.com/question/403870382/answer/1316762694 编辑:深度学习与计算机视觉声明:仅做学术分享,侵删经常看到一些人说: & ...
DL框架：主流深度学习框架(TensorFlow/Pytorch/Caffe/Keras/CNTK/MXNet/Theano/PaddlePaddle)简介、多个方向比较、案例应用之详细攻略
DL框架:主流深度学习框架(TensorFlow/Pytorch/Caffe/Keras/CNTK/MXNet/Theano/PaddlePaddle)简介.多个方向比较.案例应用之详细攻略目录深 ...
linux服务器非root用户安装Anaconda，并配置tensorflow/pytorch
文章目录 Anaconda Tensorflow Pytorch Tips Anaconda 1.清华大学镜像网站内下载需要的anaconda版本,我下载的是Anaconda3-5.3.0-Linux ...
12大深度学习开源框架（caffe,tensorflow,pytorch,mxnet等）汇总详解
这是一篇总结文,给大家来捋清楚12大深度学习开源框架的快速入门,这是有三AI的GitHub项目,欢迎大家star/fork. https://github.com/longpeng2008/yousa ...
Hinton NIPS2017 手写识别实现 TensorFlow \ Pytorch \ Keras
10月26日,深度学习元老Hinton的NIPS2017 Capsule论文<Dynamic Routing Between Capsules>终于在arxiv上发表.今天相关关于这篇论文 ...
centos GPU tensorflow pytorch 深度学习环境搭建
简要记录一下深度学习开发环境的搭建步骤,懒得编辑了,将就着吧. 总体流程参考文章:https://www.cnblogs.com/learn-the-hard-way/p/12318980.html ...
Android NNAPI - Paddle - TensorFlow - PyTorch ArgMax and ArgMin 的定义与计算过程
Android NNAPI - Paddle - TensorFlow - PyTorch ArgMax and ArgMin 的定义与计算过程 1. Android NNAPI Android ND ...
转：oracle ebs po模块一揽子采购协议小结
转自:http://yedward.net/?id=193 oracle ebs po模块一揽子采购协议小结本文总结oracle ebs 采购订单(po)模块一揽子采购协议的相关知识,总结如下: 1 ...
3090显卡安装tensorflow/pytorch
3090显卡安装tensorflow/pytorch 一 3090显卡驱动二安装CUDA11.0 三安装cuDNN 四 anaconda使用 1 安装anaconda 2 conda使用五安 ...

PSP模块Tensorflow/Pytorch实现小结