每日Attention学习1——Parallel Aggregation Pyramid Pooling Module

xiongxyowo

已于 2024-12-03 15:21:38 修改

阅读量534

点赞数 7

分类专栏：每日Attention学习文章标签：论文阅读

于 2024-05-08 23:41:38 首次发布

本文链接：https://blog.csdn.net/qq_40714949/article/details/138538710

版权

模块出处

[CVPR 23] [link] [code] PIDNet: A Real-time Semantic Segmentation Network Inspired by PID Controllers

模块名称

Parallel Aggregation Pyramid Pooling Module (PAPPM)

模块作用

多尺度特征提取，更大感受野

模块结构

在这里插入图片描述

模块代码

import torch
import torch.nn as nn
import torch.nn.functional as F

class PAPPM(nn.Module):
    def __init__(self, inplanes, branch_planes, outplanes, BatchNorm=nn.BatchNorm2d):
        super(PAPPM, self).__init__()
        bn_mom = 0.1
        
        # Avg 5,2 + Conv
        self.scale1 = nn.Sequential(nn.AvgPool2d(kernel_size=5, stride=2, padding=2),
                                    BatchNorm(inplanes, momentum=bn_mom),
                                    nn.ReLU(inplace=True),
                                    nn.Conv2d(inplanes, branch_planes, kernel_size=1, bias=False),
                                    )
        # Avg 9,4 + Conv
        self.scale2 = nn.Sequential(nn.AvgPool2d(kernel_size=9, stride=4, padding