IllTr.py

import torch
import torch.nn as nn
from torch.functional import Tensor
from torch.nn.modules.activation import Tanhshrink
from timm.models.layers import trunc_normal_
from functools import partial


class Ffn(nn.Module):
    # feed forward network layer after attention
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        x = self.drop(x)
        return x


class Attention(nn.Module):
    def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim ** -0.5

        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

    def forward(self, x, task_embed=None, level=0):
        N, L, D = x.shape
        qkv = self.qkv(x).reshape(N, L, 3, self.num_heads, D // self.num_heads).permute(2, 0, 3, 1, 4)
        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)

        # for decoder's task_embedding of different levels of attention layers
        if task_embed != None:
            _N, _H, _L, _D = q.shape
            task_embed = task_embed.reshape(1, _H, _L, _D)
            if level == 1:
                q += task_embed
                k += task_embed
            if level == 2:
                q += task_embed

        attn = (q @ k.transpose(-2, -1)) * self.scale
        attn = attn.softmax(dim=-1)
        attn = self.attn_drop(attn)

        x = (attn @ v).transpose(1, 2).reshape(N, L, D)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x


class EncoderLayer(nn.Module):
    def __init__(self, dim, num_heads, ffn_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
                 act_layer=nn.GELU, norm_layer=nn.LayerNorm):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.attn = Attention(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.norm2 = norm_layer(dim)
        ffn_hidden_dim = int(dim * ffn_ratio)
        self.ffn = Ffn(in_features=dim, hidden_features=ffn_hidden_dim, act_layer=act_layer, drop=drop)

    def forward(self, x):
        x = x + self.attn(self.norm1(x))
        x = x + self.ffn(self.norm2(x))
        return x


class DecoderLayer(nn.Module):
    def __init__(self, dim, num_heads, ffn_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
                 act_layer=nn.GELU, norm_layer=nn.LayerNorm):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.attn1 = Attention(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.norm2 = norm_layer(dim)
        self.attn2 = Attention(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.norm3 = norm_layer(dim)
        ffn_hidden_dim = int(dim * ffn_ratio)
        self.ffn = Ffn(in_features=dim, hidden_features=ffn_hidden_dim, act_layer=act_layer, drop=drop)

    def forward(self, x, task_embed):
        x = x + self.attn1(self.norm1(x), task_embed=task_embed, level=1)
        x = x + self.attn2(self.norm2(x), task_embed=task_embed, level=2)
        x = x + self.ffn(self.norm3(x))
        return x


class ResBlock(nn.Module):
    def __init__(self, channels):
        super(ResBlock, self).__init__()
        self.conv1 = nn.Conv2d(channels, channels, kernel_size=5, stride=1,
                               padding=2, bias=False)
        self.bn1 = nn.InstanceNorm2d(channels)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = nn.Conv2d(channels, channels, kernel_size=5, stride=1,
                               padding=2, bias=False)
        self.bn2 = nn.InstanceNorm2d(channels)

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        out += residual
        out = self.relu(out)

        return out


class Head(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(Head, self).__init__()
        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
                               padding=1, bias=False)
        self.bn1 = nn.InstanceNorm2d(out_channels)
        self.relu = nn.ReLU(inplace=True)
        self.resblock = ResBlock(out_channels)

    def forward(self, x):
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.resblock(out)

        return out


class PatchEmbed(nn.Module):
    """ Feature to Patch Embedding
    input : N C H W
    output: N num_patch P^2*C
    """
    def __init__(self, patch_size=1, in_channels=64):
        super().__init__()
        self.patch_size = patch_size
        self.dim = self.patch_size ** 2 * in_channels

    def forward(self, x):
        N, C, H, W = ori_shape = x.shape

        p = self.patch_size
        num_patches = (H // p) * (W // p)
        out = torch.zeros((N, num_patches, self.dim)).to(x.device)
        i, j = 0, 0
        for k in range(num_patches):
            if i + p > W:
                i = 0
                j += p
            out[:, k, :] = x[:, :, i:i + p, j:j + p].flatten(1)
            i += p
        return out, ori_shape


class DePatchEmbed(nn.Module):
    """ Patch Embedding to Feature
    input : N num_patch P^2*C
    output: N C H W
    """
    def __init__(self, patch_size=1, in_channels=64):
        super().__init__()
        self.patch_size = patch_size
        self.num_patches = None
        self.dim = self.patch_size ** 2 * in_channels

    def forward(self, x, ori_shape):
        N, num_patches, dim = x.shape
        _, C, H, W = ori_shape
        p = self.patch_size
        out = torch.zeros(ori_shape).to(x.device)
        i, j = 0, 0
        for k in range(num_patches):
            if i + p > W:
                i = 0
                j += p
            out[:, :, i:i + p, j:j + p] = x[:, k, :].reshape(N, C, p, p)
            i += p
        return out


class Tail(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(Tail, self).__init__()
        self.output = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1, bias=False)

    def forward(self, x):
        out = self.output(x)
        return out


class IllTr_Net(nn.Module):
    """ Vision Transformer with support for patch or hybrid CNN input stage
    """

    def __init__(self, patch_size=1, in_channels=3, mid_channels=16, num_classes=1000, depth=12,
                 num_heads=8, ffn_ratio=4., qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
                 norm_layer=nn.LayerNorm):
        super(IllTr_Net, self).__init__()

        self.num_classes = num_classes
        self.embed_dim = patch_size * patch_size * mid_channels
        self.head = Head(in_channels, mid_channels)
        self.patch_embedding = PatchEmbed(patch_size=patch_size, in_channels=mid_channels)
        self.embed_dim = self.patch_embedding.dim
        if self.embed_dim % num_heads != 0:
            raise RuntimeError("Embedding dim must be devided by numbers of heads")

        self.pos_embed = nn.Parameter(torch.zeros(1, (128 // patch_size) ** 2, self.embed_dim))
        self.task_embed = nn.Parameter(torch.zeros(6, 1, (128 // patch_size) ** 2, self.embed_dim))

        self.encoder = nn.ModuleList([
            EncoderLayer(
                dim=self.embed_dim, num_heads=num_heads, ffn_ratio=ffn_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                drop=drop_rate, attn_drop=attn_drop_rate, norm_layer=norm_layer)
            for _ in range(depth)])
        self.decoder = nn.ModuleList([
            DecoderLayer(
                dim=self.embed_dim, num_heads=num_heads, ffn_ratio=ffn_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                drop=drop_rate, attn_drop=attn_drop_rate, norm_layer=norm_layer)
            for _ in range(depth)])

        self.de_patch_embedding = DePatchEmbed(patch_size=patch_size, in_channels=mid_channels)
        # tail
        self.tail = Tail(int(mid_channels), in_channels)
        
        self.acf = nn.Hardtanh(0,1)

        trunc_normal_(self.pos_embed, std=.02)
        self.apply(self._init_weights)

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            trunc_normal_(m.weight, std=.02)
            if isinstance(m, nn.Linear) and m.bias is not None:
                nn.init.constant_(m.bias, 0)
        elif isinstance(m, nn.LayerNorm):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    def forward(self, x):
        x = self.head(x)
        x, ori_shape = self.patch_embedding(x)
        x = x + self.pos_embed[:, :x.shape[1]]
        
        for blk in self.encoder:
            x = blk(x)

        for blk in self.decoder:
            x = blk(x, self.task_embed[0, :, :x.shape[1]])
      
        x = self.de_patch_embedding(x, ori_shape)
        x = self.tail(x)
        
        x = self.acf(x)
        return x


def IllTr(**kwargs):
    model = IllTr_Net(
        patch_size=4, depth=6, num_heads=8, ffn_ratio=4, qkv_bias=True, norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs)

    return model