import torch import torch.nn as nn import torch.nn.functional as F from functools import partial import math from .helpers import load_pretrained from .layers import DropPath, to_2tuple, trunc_normal_ from ..builder import HEADS from .decode_head import BaseDecodeHead from ..backbones.vit import Block from mmcv.cnn import build_norm_layer class MLAHead(nn.Module): def __init__(self, mla_channels=256, mlahead_channels=128, norm_cfg=None): super(MLAHead, self).__init__() self.head2 = nn.Sequential(nn.Conv2d(mla_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU(), nn.Conv2d(mlahead_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU()) self.head3 = nn.Sequential(nn.Conv2d(mla_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU(), nn.Conv2d(mlahead_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU()) self.head4 = nn.Sequential(nn.Conv2d(mla_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU(), nn.Conv2d(mlahead_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU()) self.head5 = nn.Sequential(nn.Conv2d(mla_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU(), nn.Conv2d(mlahead_channels, mlahead_channels, 3, padding=1, bias=False), build_norm_layer(norm_cfg, mlahead_channels)[1], nn.ReLU()) def forward(self, mla_p2, mla_p3, mla_p4, mla_p5): # head2 = self.head2(mla_p2) head2 = F.interpolate(self.head2(mla_p2), 4*mla_p2.shape[-1], mode='bilinear', align_corners=True) head3 = F.interpolate(self.head3(mla_p3), 4*mla_p3.shape[-1], mode='bilinear', align_corners=True) head4 = F.interpolate(self.head4(mla_p4), 4*mla_p4.shape[-1], mode='bilinear', align_corners=True) head5 = F.interpolate(self.head5(mla_p5), 4*mla_p5.shape[-1], mode='bilinear', align_corners=True) return torch.cat([head2, head3, head4, head5], dim=1) @HEADS.register_module() class VIT_MLAHead(BaseDecodeHead): """ Vision Transformer with support for patch or hybrid CNN input stage """ def __init__(self, img_size=768, mla_channels=256, mlahead_channels=128, norm_layer=nn.BatchNorm2d, norm_cfg=None, **kwargs): super(VIT_MLAHead, self).__init__(**kwargs) self.img_size = img_size self.norm_cfg = norm_cfg self.mla_channels = mla_channels self.BatchNorm = norm_layer self.mlahead_channels = mlahead_channels self.mlahead = MLAHead(mla_channels=self.mla_channels, mlahead_channels=self.mlahead_channels, norm_cfg=self.norm_cfg) self.cls = nn.Conv2d(4 * self.mlahead_channels, self.num_classes, 3, padding=1) def forward(self, inputs): x = self.mlahead(inputs[0], inputs[1], inputs[2], inputs[3]) x = self.cls(x) x = F.interpolate(x, size=self.img_size, mode='bilinear', align_corners=self.align_corners) return x