import torch import numpy as np import torch.nn as nn from timm.models.layers import DropPath, LayerNorm2d def window_partition(x, window_size): B, C, H, W = x.shape x = x.view(B, C, H // window_size, window_size, W // window_size, window_size) windows = x.permute(0, 2, 4, 3, 5, 1).reshape(-1, window_size*window_size, C) return windows def window_reverse(windows, window_size, H, W, B): x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) x = x.permute(0, 5, 1, 3, 2, 4).reshape(B, windows.shape[2], H, W) return x def ct_dewindow(ct, W, H, window_size): bs = ct.shape[0] N=ct.shape[2] ct2 = ct.view(-1, W//window_size, H//window_size, window_size, window_size, N).permute(0, 5, 1, 3, 2, 4) ct2 = ct2.reshape(bs, N, W*H).transpose(1, 2) return ct2 def ct_window(ct, W, H, window_size): bs = ct.shape[0] N = ct.shape[2] ct = ct.view(bs, H // window_size, window_size, W // window_size, window_size, N) ct = ct.permute(0, 1, 3, 2, 4, 5) return ct class PosEmbMLPSwinv2D(nn.Module): def __init__(self, window_size, pretrained_window_size, num_heads, seq_length, ct_correct=False, no_log=False): super().__init__() self.window_size = window_size self.num_heads = num_heads self.cpb_mlp = nn.Sequential(nn.Linear(2, 512, bias=True), nn.ReLU(inplace=True), nn.Linear(512, num_heads, bias=False)) relative_coords_h = torch.arange(-(self.window_size[0] - 1), self.window_size[0], dtype=torch.float32) relative_coords_w = torch.arange(-(self.window_size[1] - 1), self.window_size[1], dtype=torch.float32) relative_coords_table = torch.stack( torch.meshgrid([relative_coords_h, relative_coords_w])).permute(1, 2, 0).contiguous().unsqueeze(0) # 1, 2*Wh-1, 2*Ww-1, 2 if pretrained_window_size[0] > 0: relative_coords_table[:, :, :, 0] /= (pretrained_window_size[0] - 1) relative_coords_table[:, :, :, 1] /= (pretrained_window_size[1] - 1) else: relative_coords_table[:, :, :, 0] /= (self.window_size[0] - 1) relative_coords_table[:, :, :, 1] /= (self.window_size[1] - 1) if not no_log: relative_coords_table *= 8 # normalize to -8, 8 relative_coords_table = torch.sign(relative_coords_table) * torch.log2( torch.abs(relative_coords_table) + 1.0) / np.log2(8) self.register_buffer("relative_coords_table", relative_coords_table) coords_h = torch.arange(self.window_size[0]) coords_w = torch.arange(self.window_size[1]) coords = torch.stack(torch.meshgrid([coords_h, coords_w])) coords_flatten = torch.flatten(coords, 1) relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :] relative_coords = relative_coords.permute(1, 2, 0).contiguous() relative_coords[:, :, 0] += self.window_size[0] - 1 relative_coords[:, :, 1] += self.window_size[1] - 1 relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1 relative_position_index = relative_coords.sum(-1) self.register_buffer("relative_position_index", relative_position_index) self.grid_exists = False self.pos_emb = None self.deploy = False relative_bias = torch.zeros(1, num_heads, seq_length, seq_length) self.seq_length = seq_length self.register_buffer("relative_bias", relative_bias) self.ct_correct=ct_correct def switch_to_deploy(self): self.deploy = True def forward(self, input_tensor, local_window_size): if self.deploy: input_tensor += self.relative_bias return input_tensor else: self.grid_exists = False if not self.grid_exists: self.grid_exists = True relative_position_bias_table = self.cpb_mlp(self.relative_coords_table).view(-1, self.num_heads) relative_position_bias = relative_position_bias_table[self.relative_position_index.view(-1)].view( self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1) relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() relative_position_bias = 16 * torch.sigmoid(relative_position_bias) n_global_feature = input_tensor.shape[2] - local_window_size if n_global_feature > 0 and self.ct_correct: step_for_ct=self.window_size[0]/(n_global_feature**0.5+1) seq_length = int(n_global_feature ** 0.5) indices = [] for i in range(seq_length): for j in range(seq_length): ind = (i+1)*step_for_ct*self.window_size[0] + (j+1)*step_for_ct indices.append(int(ind)) top_part = relative_position_bias[:, indices, :] lefttop_part = relative_position_bias[:, indices, :][:, :, indices] left_part = relative_position_bias[:, :, indices] relative_position_bias = torch.nn.functional.pad(relative_position_bias, (n_global_feature, 0, n_global_feature, 0)).contiguous() if n_global_feature>0 and self.ct_correct: relative_position_bias = relative_position_bias*0.0 relative_position_bias[:, :n_global_feature, :n_global_feature] = lefttop_part relative_position_bias[:, :n_global_feature, n_global_feature:] = top_part relative_position_bias[:, n_global_feature:, :n_global_feature] = left_part self.pos_emb = relative_position_bias.unsqueeze(0) self.relative_bias = self.pos_emb input_tensor += self.pos_emb return input_tensor class PosEmbMLPSwinv1D(nn.Module): def __init__(self, dim, rank=2, seq_length=4, conv=False): super().__init__() self.rank = rank if not conv: self.cpb_mlp = nn.Sequential(nn.Linear(self.rank, 512, bias=True), nn.ReLU(), nn.Linear(512, dim, bias=False)) else: self.cpb_mlp = nn.Sequential(nn.Conv1d(self.rank, 512, 1,bias=True), nn.ReLU(), nn.Conv1d(512, dim, 1,bias=False)) self.grid_exists = False self.pos_emb = None self.deploy = False relative_bias = torch.zeros(1,seq_length, dim) self.register_buffer("relative_bias", relative_bias) self.conv = conv def switch_to_deploy(self): self.deploy = True def forward(self, input_tensor): seq_length = input_tensor.shape[1] if not self.conv else input_tensor.shape[2] if self.deploy: return input_tensor + self.relative_bias else: self.grid_exists = False if not self.grid_exists: self.grid_exists = True if self.rank == 1: relative_coords_h = torch.arange(0, seq_length, device=input_tensor.device, dtype = input_tensor.dtype) relative_coords_h -= seq_length//2 relative_coords_h /= (seq_length//2) relative_coords_table = relative_coords_h self.pos_emb = self.cpb_mlp(relative_coords_table.unsqueeze(0).unsqueeze(2)) self.relative_bias = self.pos_emb else: seq_length = int(seq_length**0.5) relative_coords_h = torch.arange(0, seq_length, device=input_tensor.device, dtype = input_tensor.dtype) relative_coords_w = torch.arange(0, seq_length, device=input_tensor.device, dtype = input_tensor.dtype) relative_coords_table = torch.stack(torch.meshgrid([relative_coords_h, relative_coords_w])).contiguous().unsqueeze(0) relative_coords_table -= seq_length // 2 relative_coords_table /= (seq_length // 2) if not self.conv: self.pos_emb = self.cpb_mlp(relative_coords_table.flatten(2).transpose(1,2)) else: self.pos_emb = self.cpb_mlp(relative_coords_table.flatten(2)) self.relative_bias = self.pos_emb input_tensor = input_tensor + self.pos_emb return input_tensor class Mlp(nn.Module): """ Multi-Layer Perceptron (MLP) block """ def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.): """ Args: in_features: input features dimension. hidden_features: hidden features dimension. out_features: output features dimension. act_layer: activation function. drop: dropout rate. """ 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_size = x.size() x = x.view(-1, x_size[-1]) x = self.fc1(x) x = self.act(x) x = self.drop(x) x = self.fc2(x) x = self.drop(x) x = x.view(x_size) return x class Downsample(nn.Module): """ Down-sampling block based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, dim, out_dim, keep_dim=False, stride=2, ): """ Args: dim: feature size dimension. norm_layer: normalization layer. keep_dim: bool argument for maintaining the resolution. """ super().__init__() if keep_dim: out_dim = dim self.norm = LayerNorm2d(dim) self.reduction = nn.Sequential( nn.Conv2d(dim, out_dim, 3, stride, 1, bias=False), ) def forward(self, x): x = self.norm(x) x = self.reduction(x) return x class PatchEmbed(nn.Module): """ Patch embedding block based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, in_chans=3, in_dim=64, dim=96): """ Args: in_chans: number of input channels. dim: feature size dimension. """ super().__init__() self.proj = nn.Identity() self.conv_down = nn.Sequential( nn.Conv2d(in_chans, in_dim, 3, 2, 1, bias=False), nn.BatchNorm2d(in_dim, eps=1e-4), nn.ReLU(), nn.Conv2d(in_dim, dim, 3, 2, 1, bias=False), nn.BatchNorm2d(dim, eps=1e-4), nn.ReLU() ) def forward(self, x): x = self.proj(x) x = self.conv_down(x) return x class ConvBlock(nn.Module): """ Conv block based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, dim, drop_path=0., layer_scale=None, kernel_size=3): super().__init__() """ Args: drop_path: drop path. layer_scale: layer scale coefficient. kernel_size: kernel size. """ self.conv1 = nn.Conv2d(dim, dim, kernel_size=kernel_size, stride=1, padding=1) self.norm1 = nn.BatchNorm2d(dim, eps=1e-5) self.act1 = nn.GELU() self.conv2 = nn.Conv2d(dim, dim, kernel_size=kernel_size, stride=1, padding=1) self.norm2 = nn.BatchNorm2d(dim, eps=1e-5) self.layer_scale = layer_scale if layer_scale is not None and type(layer_scale) in [int, float]: self.gamma = nn.Parameter(layer_scale * torch.ones(dim)) self.layer_scale = True else: self.layer_scale = False self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() def forward(self, x, global_feature=None): input = x x = self.conv1(x) x = self.norm1(x) x = self.act1(x) x = self.conv2(x) x = self.norm2(x) if self.layer_scale: x = x * self.gamma.view(1, -1, 1, 1) x = input + self.drop_path(x) return x, global_feature class WindowAttention(nn.Module): """ Window attention based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0., resolution=0, seq_length=0): super().__init__() """ Args: dim: feature size dimension. num_heads: number of attention head. qkv_bias: bool argument for query, key, value learnable bias. qk_scale: bool argument to scaling query, key. attn_drop: attention dropout rate. proj_drop: output dropout rate. resolution: feature resolution. seq_length: sequence length. """ self.num_heads = num_heads head_dim = dim // num_heads self.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) # attention positional bias self.pos_emb_funct = PosEmbMLPSwinv2D(window_size=[resolution, resolution], pretrained_window_size=[resolution, resolution], num_heads=num_heads, seq_length=seq_length) self.resolution = resolution def forward(self, x): B, N, C = x.shape qkv = self.qkv(x).reshape(B, -1, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) q, k, v = qkv[0], qkv[1], qkv[2] attn = (q @ k.transpose(-2, -1)) * self.scale attn = self.pos_emb_funct(attn, self.resolution ** 2) attn = attn.softmax(dim=-1) attn = self.attn_drop(attn) x = (attn @ v).transpose(1, 2).reshape(B, -1, C) x = self.proj(x) x = self.proj_drop(x) return x class HAT(nn.Module): """ Hierarchical attention (HAT) based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0., drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, sr_ratio=1., window_size=7, last=False, layer_scale=None, ct_size=1, do_propagation=False): super().__init__() """ Args: dim: feature size dimension. num_heads: number of attention head. mlp_ratio: MLP ratio. qkv_bias: bool argument for query, key, value learnable bias. qk_scale: bool argument to scaling query, key. drop: dropout rate. attn_drop: attention dropout rate. proj_drop: output dropout rate. act_layer: activation function. norm_layer: normalization layer. sr_ratio: input to window size ratio. window_size: window size. last: last layer flag. layer_scale: layer scale coefficient. ct_size: spatial dimension of carrier token local window. do_propagation: enable carrier token propagation. """ # positional encoding for windowed attention tokens self.pos_embed = PosEmbMLPSwinv1D(dim, rank=2, seq_length=window_size**2) self.norm1 = norm_layer(dim) # number of carrier tokens per every window cr_tokens_per_window = ct_size**2 if sr_ratio > 1 else 0 # total number of carrier tokens cr_tokens_total = cr_tokens_per_window*sr_ratio*sr_ratio self.cr_window = ct_size self.attn = WindowAttention(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop, resolution=window_size, seq_length=window_size**2 + cr_tokens_per_window) self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() self.norm2 = norm_layer(dim) mlp_hidden_dim = int(dim * mlp_ratio) self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) self.window_size = window_size use_layer_scale = layer_scale is not None and type(layer_scale) in [int, float] self.gamma3 = nn.Parameter(layer_scale * torch.ones(dim)) if use_layer_scale else 1 self.gamma4 = nn.Parameter(layer_scale * torch.ones(dim)) if use_layer_scale else 1 self.sr_ratio = sr_ratio if sr_ratio > 1: # if do hierarchical attention, this part is for carrier tokens self.hat_norm1 = norm_layer(dim) self.hat_norm2 = norm_layer(dim) self.hat_attn = WindowAttention( dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop, resolution=int(cr_tokens_total**0.5), seq_length=cr_tokens_total) self.hat_mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) self.hat_drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() self.hat_pos_embed = PosEmbMLPSwinv1D(dim, rank=2, seq_length=cr_tokens_total) self.gamma1 = nn.Parameter(layer_scale * torch.ones(dim)) if use_layer_scale else 1 self.gamma2 = nn.Parameter(layer_scale * torch.ones(dim)) if use_layer_scale else 1 self.upsampler = nn.Upsample(size=window_size, mode='nearest') # keep track for the last block to explicitly add carrier tokens to feature maps self.last = last self.do_propagation = do_propagation def forward(self, x, carrier_tokens): B, T, N = x.shape ct = carrier_tokens x = self.pos_embed(x) if self.sr_ratio > 1: # do hierarchical attention via carrier tokens # first do attention for carrier tokens Bg, Ng, Hg = ct.shape # ct are located quite differently ct = ct_dewindow(ct, self.cr_window*self.sr_ratio, self.cr_window*self.sr_ratio, self.cr_window) # positional bias for carrier tokens ct = self.hat_pos_embed(ct) # attention plus mlp ct = ct + self.hat_drop_path(self.gamma1*self.hat_attn(self.hat_norm1(ct))) ct = ct + self.hat_drop_path(self.gamma2*self.hat_mlp(self.hat_norm2(ct))) # ct are put back to windows ct = ct_window(ct, self.cr_window * self.sr_ratio, self.cr_window * self.sr_ratio, self.cr_window) ct = ct.reshape(x.shape[0], -1, N) # concatenate carrier_tokens to the windowed tokens x = torch.cat((ct, x), dim=1) # window attention together with carrier tokens x = x + self.drop_path(self.gamma3*self.attn(self.norm1(x))) x = x + self.drop_path(self.gamma4*self.mlp(self.norm2(x))) if self.sr_ratio > 1: # for hierarchical attention we need to split carrier tokens and window tokens back ctr, x = x.split([x.shape[1] - self.window_size*self.window_size, self.window_size*self.window_size], dim=1) ct = ctr.reshape(Bg, Ng, Hg) # reshape carrier tokens. if self.last and self.do_propagation: # propagate carrier token information into the image ctr_image_space = ctr.transpose(1, 2).reshape(B, N, self.cr_window, self.cr_window) x = x + self.gamma1 * self.upsampler(ctr_image_space.to(dtype=torch.float32)).flatten(2).transpose(1, 2).to(dtype=x.dtype) return x, ct class TokenInitializer(nn.Module): """ Carrier token Initializer based on: "Hatamizadeh et al., FasterViT: Fast Vision Transformers with Hierarchical Attention """ def __init__(self, dim, input_resolution, window_size, ct_size=1): """ Args: dim: feature size dimension. input_resolution: input image resolution. window_size: window size. ct_size: spatial dimension of carrier token local window """ super().__init__() output_size = int(ct_size * input_resolution/window_size) stride_size = int(input_resolution/output_size) kernel_size = input_resolution - (output_size - 1) * stride_size self.pos_embed = nn.Conv2d(dim, dim, 3, padding=1, groups=dim) to_global_feature = nn.Sequential() to_global_feature.add_module("pos", self.pos_embed) to_global_feature.add_module("pool", nn.AvgPool2d(kernel_size=kernel_size, stride=stride_size)) self.to_global_feature = to_global_feature self.window_size = ct_size def forward(self, x): x = self.to_global_feature(x) B, C, H, W = x.shape ct = x.view(B, C, H // self.window_size, self.window_size, W // self.window_size, self.window_size) ct = ct.permute(0, 2, 4, 3, 5, 1).reshape(-1, H*W, C) return ct class FasterViTLayer(nn.Module): """ GCViT layer based on: "Hatamizadeh et al., Global Context Vision Transformers " """ def __init__(self, dim, out_dim, depth, input_resolution, num_heads, window_size, ct_size=1, conv=False, downsample=True, mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0., layer_scale=None, layer_scale_conv=None, only_local=False, hierarchy=True, do_propagation=False ): """ Args: dim: feature size dimension. depth: layer depth. input_resolution: input resolution. num_heads: number of attention head. window_size: window size. ct_size: spatial dimension of carrier token local window. conv: conv_based stage flag. downsample: downsample flag. mlp_ratio: MLP ratio. qkv_bias: bool argument for query, key, value learnable bias. qk_scale: bool argument to scaling query, key. drop: dropout rate. attn_drop: attention dropout rate. drop_path: drop path rate. layer_scale: layer scale coefficient. layer_scale_conv: conv layer scale coefficient. only_local: local attention flag. hierarchy: hierarchical attention flag. do_propagation: enable carrier token propagation. """ super().__init__() self.conv = conv self.transformer_block = False if conv: self.blocks = nn.ModuleList([ ConvBlock(dim=dim, drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path, layer_scale=layer_scale_conv) for i in range(depth)]) self.transformer_block = False else: sr_ratio = input_resolution // window_size if not only_local else 1 self.blocks = nn.ModuleList([ HAT(dim=dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale, drop=drop, attn_drop=attn_drop, drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path, sr_ratio=sr_ratio, window_size=window_size, last=(i == depth-1), layer_scale=layer_scale, ct_size=ct_size, do_propagation=do_propagation, ) for i in range(depth)]) self.transformer_block = True self.downsample = Downsample(dim=dim, out_dim=out_dim, stride=1) if not downsample else Downsample(dim=dim, out_dim=out_dim, stride=2) if len(self.blocks) and not only_local and input_resolution // window_size > 1 and hierarchy and not self.conv: self.global_tokenizer = TokenInitializer(dim, input_resolution, window_size, ct_size=ct_size) self.do_gt = True else: self.do_gt = False self.window_size = window_size def forward(self, x): ct = self.global_tokenizer(x) if self.do_gt else None B, C, H, W = x.shape if self.transformer_block: x = window_partition(x, self.window_size) for bn, blk in enumerate(self.blocks): x, ct = blk(x, ct) if self.transformer_block: x = window_reverse(x, self.window_size, H, W, B) if self.downsample is None: return x return self.downsample(x)