# code was heavily based on https://github.com/wtjiang98/PSGAN # MIT License # Copyright (c) 2020 Wentao Jiang import paddle import paddle.nn as nn import paddle.nn.functional as F from paddle.vision.models import resnet18 class ConvBNReLU(paddle.nn.Layer): def __init__(self, in_chan, out_chan, ks=3, stride=1, padding=1, *args, **kwargs): super(ConvBNReLU, self).__init__() self.conv = nn.Conv2D( in_chan, out_chan, kernel_size=ks, stride=stride, padding=padding, bias_attr=False) self.bn = nn.BatchNorm2D(out_chan) self.relu = nn.ReLU() def forward(self, x): x = self.conv(x) x = self.bn(x) x = self.relu(x) return x class BiSeNetOutput(paddle.nn.Layer): def __init__(self, in_chan, mid_chan, n_classes, *args, **kwargs): super(BiSeNetOutput, self).__init__() self.conv = ConvBNReLU(in_chan, mid_chan, ks=3, stride=1, padding=1) self.conv_out = nn.Conv2D( mid_chan, n_classes, kernel_size=1, bias_attr=False) def forward(self, x): x = self.conv(x) x = self.conv_out(x) return x class AttentionRefinementModule(paddle.nn.Layer): def __init__(self, in_chan, out_chan, *args, **kwargs): super(AttentionRefinementModule, self).__init__() self.conv = ConvBNReLU(in_chan, out_chan, ks=3, stride=1, padding=1) self.conv_atten = nn.Conv2D( out_chan, out_chan, kernel_size=1, bias_attr=False) self.bn_atten = nn.BatchNorm(out_chan) self.sigmoid_atten = nn.Sigmoid() def forward(self, x): feat = self.conv(x) atten = F.avg_pool2d(feat, feat.shape[2:]) atten = self.conv_atten(atten) atten = self.bn_atten(atten) atten = self.sigmoid_atten(atten) out = feat * atten return out class ContextPath(paddle.nn.Layer): def __init__(self, *args, **kwargs): super(ContextPath, self).__init__() self.backbone = resnet18(pretrained=True) self.arm16 = AttentionRefinementModule(256, 128) self.arm32 = AttentionRefinementModule(512, 128) self.conv_head32 = ConvBNReLU(128, 128, ks=3, stride=1, padding=1) self.conv_head16 = ConvBNReLU(128, 128, ks=3, stride=1, padding=1) self.conv_avg = ConvBNReLU(512, 128, ks=1, stride=1, padding=0) def backbone_forward(self, x): x = self.backbone.conv1(x) x = self.backbone.bn1(x) x = self.backbone.relu(x) x = self.backbone.maxpool(x) x = self.backbone.layer1(x) c2 = self.backbone.layer2(x) c3 = self.backbone.layer3(c2) c4 = self.backbone.layer4(c3) return c2, c3, c4 def forward(self, x): H0, W0 = x.shape[2:] feat8, feat16, feat32 = self.backbone_forward(x) H8, W8 = feat8.shape[2:] H16, W16 = feat16.shape[2:] H32, W32 = feat32.shape[2:] avg = F.avg_pool2d(feat32, feat32.shape[2:]) avg = self.conv_avg(avg) avg_up = F.interpolate(avg, size=(H32, W32), mode='nearest') feat32_arm = self.arm32(feat32) feat32_sum = feat32_arm + avg_up feat32_up = F.interpolate(feat32_sum, size=(H16, W16), mode='nearest') feat32_up = self.conv_head32(feat32_up) feat16_arm = self.arm16(feat16) feat16_sum = feat16_arm + feat32_up feat16_up = F.interpolate(feat16_sum, size=(H8, W8), mode='nearest') feat16_up = self.conv_head16(feat16_up) return feat8, feat16_up, feat32_up # x8, x8, x16 class SpatialPath(paddle.nn.Layer): def __init__(self, *args, **kwargs): super(SpatialPath, self).__init__() self.conv1 = ConvBNReLU(3, 64, ks=7, stride=2, padding=3) self.conv2 = ConvBNReLU(64, 64, ks=3, stride=2, padding=1) self.conv3 = ConvBNReLU(64, 64, ks=3, stride=2, padding=1) self.conv_out = ConvBNReLU(64, 128, ks=1, stride=1, padding=0) def forward(self, x): feat = self.conv1(x) feat = self.conv2(feat) feat = self.conv3(feat) feat = self.conv_out(feat) return feat class FeatureFusionModule(paddle.nn.Layer): def __init__(self, in_chan, out_chan, *args, **kwargs): super(FeatureFusionModule, self).__init__() self.convblk = ConvBNReLU(in_chan, out_chan, ks=1, stride=1, padding=0) self.conv1 = nn.Conv2D( out_chan, out_chan // 4, kernel_size=1, stride=1, padding=0, bias_attr=False) self.conv2 = nn.Conv2D( out_chan // 4, out_chan, kernel_size=1, stride=1, padding=0, bias_attr=False) self.relu = nn.ReLU() self.sigmoid = nn.Sigmoid() def forward(self, fsp, fcp): fcat = paddle.concat([fsp, fcp], axis=1) feat = self.convblk(fcat) atten = F.avg_pool2d(feat, feat.shape[2:]) atten = self.conv1(atten) atten = self.relu(atten) atten = self.conv2(atten) atten = self.sigmoid(atten) feat_atten = feat * atten feat_out = feat_atten + feat return feat_out class BiSeNet(paddle.nn.Layer): def __init__(self, n_classes, *args, **kwargs): super(BiSeNet, self).__init__() self.cp = ContextPath() self.ffm = FeatureFusionModule(256, 256) self.conv_out = BiSeNetOutput(256, 256, n_classes) self.conv_out16 = BiSeNetOutput(128, 64, n_classes) self.conv_out32 = BiSeNetOutput(128, 64, n_classes) def forward(self, x): H, W = x.shape[2:] feat_res8, feat_cp8, feat_cp16 = self.cp( x) # here return res3b1 feature feat_sp = feat_res8 # use res3b1 feature to replace spatial path feature feat_fuse = self.ffm(feat_sp, feat_cp8) feat_out = self.conv_out(feat_fuse) feat_out16 = self.conv_out16(feat_cp8) feat_out32 = self.conv_out32(feat_cp16) feat_out = F.interpolate(feat_out, size=(H, W)) feat_out16 = F.interpolate(feat_out16, size=(H, W)) feat_out32 = F.interpolate(feat_out32, size=(H, W)) return feat_out, feat_out16, feat_out32