ASPP的pytorch实现

 

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

class ASPPConv(nn.Sequential):
    def __init__(self, in_channels, out_channels, dilation):
        modules = [
            nn.Conv2d(in_channels, out_channels, 3, padding=dilation, dilation=dilation, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU()
        ]
        super(ASPPConv, self).__init__(*modules)

class ASPPPooling(nn.Sequential):
    def __init__(self, in_channels, out_channels):
        super(ASPPPooling, self).__init__(
            nn.AdaptiveAvgPool2d(1),
            nn.Conv2d(in_channels, out_channels, 1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU())

    def forward(self, x):
        size = x.shape[-2:]
        x = super(ASPPPooling, self).forward(x)
        return F.interpolate(x, size=size, mode='bilinear', align_corners=False)

class ASPP(nn.Module):
    def __init__(self, in_channels, atrous_rates):
        super(ASPP, self).__init__()
        out_channels = 256
        modules = []
        modules.append(nn.Sequential(
            nn.Conv2d(in_channels, out_channels, 1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU()))

        rate1, rate2, rate3 = tuple(atrous_rates)
        modules.append(ASPPConv(in_channels, out_channels, rate1))
        modules.append(ASPPConv(in_channels, out_channels, rate2))
        modules.append(ASPPConv(in_channels, out_channels, rate3))
        modules.append(ASPPPooling(in_channels, out_channels))

        self.convs = nn.ModuleList(modules)

        self.project = nn.Sequential(
            nn.Conv2d(5 * out_channels, out_channels, 1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(),
            nn.Dropout(0.5))

    def forward(self, x):
        res = []
        for conv in self.convs:
            res.append(conv(x))
        res = torch.cat(res, dim=1)
        return self.project(res)

aspp = ASPP(256,[6,12,18])
x = torch.rand(2,256,13,13)
print(aspp(x).shape)