SwinTransformer学习记录（二）之SwinTransformer Block

彭祥.

于 2024-05-03 15:16:49 发布

阅读量799

点赞数 21

分类专栏： SwinTransformer 文章标签：学习

本文链接：https://blog.csdn.net/pengxiang1998/article/details/133783313

版权

SwinTransformer 专栏收录该内容

2 篇文章 0 订阅

订阅专栏

在前面的章节中，我们学习了SwinTransformer的整体框架，其主要由Patch Merging模块与SwinTansformer Block模块组成，

Patch Embedding

在输入进Swin Transformer Block 前，需要将图片切成一个个 patch，然后嵌入向量。

具体做法是对原始图片裁成一个个 window_size * window_size 的窗口大小，然后进行嵌入。

这里可以通过二维卷积层，将 stride，kernel_size 设置为 window_size 大小。设定输出通道来确定嵌入向量的大小。最后将 H,W 维度展开，并移动到第一维度。这里的window_size设置为4，具体过程如下，其他阶段的Patch也是如法炮制。

在这里插入图片描述

class PatchEmbed(nn.Module):
    """ Image to Patch Embedding
    Args:
        patch_size (int): Patch token size. Default: 4.
        in_chans (int): Number of input image channels. Default: 3.
        embed_dim (int): Number of linear projection output channels. Default: 96.
        norm_layer (nn.Module, optional): Normalization layer. Default: None
    """
    def __init__(self, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
        super().__init__()
        patch_size = to_2tuple(patch_size)
        self.patch_size = patch_size
        self.in_chans = in_chans
        self.embed_dim = embed_dim
        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
        if norm_layer is not None:
            self.norm = norm_layer(embed_dim)
        else:
            self.norm = None
    def forward(self, x):
        """Forward function."""
        # padding
        _, _, H, W = x.size()
        if W % self.patch_size[1] != 0:
            x = F.pad(x, (0, self.patch_size[1] - W % self.patch_size[1]))
        if H % self.patch_size[0] != 0:
            x = F.pad(x, (0, 0, 0, self.patch_size[0] - H % self.patch_size[0]))
        x = self.proj(x)  # B C Wh Ww
        if self.norm is not None:
            Wh, Ww = x.size(2), x.size(3)
            x = x.flatten(2).transpose(1, 2)
            x = self.norm(x)
            x = x.transpose(1, 2).view(-1, self.embed_dim, Wh, Ww)
        return x

Patch Merging

该模块的作用是在每个stage 开始前做降采样，用于缩小分辨率，调整通道数。

在 CNN 中，则是在每个 Stage 开始前用stride=2的卷积/池化层来降低分辨率。
每次降采样是两倍，因此在行方向和列方向上，间隔 2 选取元素。

然后拼接在一起作为一整个张量，最后展开。此时通道维度会变成原先的 4 倍（因为 H,W 各缩小 2 倍），此时再通过一个全连接层再调整通道维度为原来的两倍。如下图所示：

在这里插入图片描述

class PatchMerging(nn.Module):
    """ Patch Merging Layer
    Args:
        dim (int): Number of input channels.
        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
    """
    def __init__(self, dim, norm_layer=nn.LayerNorm):
        super().__init__()
        self.dim = dim
        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
        self.norm = norm_layer(4 * dim)
    def forward(self, x, H, W):
        B, L, C = x.shape
        assert L == H * W, "input feature has wrong size"
        x = x.view(B, H, W, C)
        # padding
        pad_input = (H % 2 == 1) or (W % 2 == 1)
        if pad_input:
            x = F.pad(x, (0, 0, 0, W % 2, 0, H % 2))
        x0 = x[:, 0::2, 0::2, :]  # B H/2 W/2 C
        x1 = x[:, 1::2, 0::2, :]  # B H/2 W/2 C
        x2 = x[:, 0::2, 1::2, :]  # B H/2 W/2 C
        x3 = x[:, 1::2, 1::2, :]  # B H/2 W/2 C
        x = torch.cat([x0, x1, x2, x3], -1)  # B H/2 W/2 4*C
        x = x.view(B, -1, 4 * C)  # B H/2*W/2 4*C
        x = self.norm(x)
        x = self.reduction(x)
        return x

接着便是进入SwinTransformer Block模块了，由于每个阶段中SwinTransformer Block模块的个数不相同，因此代码中定义BasicLayer来整合SwinTransformer Block

BasicLayer(整合SwinTransformer Block）

BasicLayer与SwinTransformer Block的区别在于：BasicLayer是多个SwinTransformer Block

class BasicLayer(nn.Module):
    """ A basic Swin Transformer layer for one stage.
    Args:
        dim (int): Number of feature channels
        depth (int): Depths of this stage.
        num_heads (int): Number of attention head.
        window_size (int): Local window size. Default: 7.
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
        drop (float, optional): Dropout rate. Default: 0.0
        attn_drop (float, optional): Attention dropout rate. Default: 0.0
        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
    """

    def __init__(self,
                 dim,
                 depth,
                 num_heads,
                 window_size=7,
                 mlp_ratio=4.,
                 qkv_bias=True,
                 qk_scale=None,
                 drop=0.,
                 attn_drop=0.,
                 drop_path=0.,
                 norm_layer=nn.LayerNorm,
                 downsample=None,
                 use_checkpoint=False):
        super().__init__()
        self.window_size = window_size
        self.shift_size = window_size // 2
        self.depth = depth
        self.use_checkpoint = use_checkpoint

        # build blocks
        self.blocks = nn.ModuleList([
            SwinTransformerBlock(
                dim=dim,
                num_heads=num_heads,
                window_size=window_size,
                shift_size=0 if (i % 2 == 0) else window_size // 2,
                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,
                norm_layer=norm_layer)
            for i in range(depth)])

        # patch merging layer
        if downsample is not None:
            self.downsample = downsample(dim=dim, norm_layer=norm_layer)
        else:
            self.downsample = None

    def forward(self, x, H, W):
        """ Forward function.
        Args:
            x: Input feature, tensor size (B, H*W, C).
            H, W: Spatial resolution of the input feature.
        """

        # calculate attention mask for SW-MSA
        Hp = int(np.ceil(H / self.window_size)) * self.window_size
        Wp = int(np.ceil(W / self.window_size)) * self.window_size
        img_mask = torch.zeros((1, Hp, Wp, 1), device=x.device)  # 1 Hp Wp 1
        h_slices = (slice(0, -self.window_size),
                    slice(-self.window_size, -self.shift_size),
                    slice(-self.shift_size, None))
        w_slices = (slice(0, -self.window_size),
                    slice(-self.window_size, -self.shift_size),
                    slice(-self.shift_size, None))
        cnt = 0
        for h in h_slices:
            for w in w_slices:
                img_mask[:, h, w, :] = cnt
                cnt += 1

        mask_windows = window_partition(img_mask, self.window_size)  # nW, window_size, window_size, 1
        mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
        attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
        attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))

        for blk in self.blocks:
            blk.H, blk.W = H, W
            if self.use_checkpoint:
                x = checkpoint.checkpoint(blk, x, attn_mask)
            else:
                x = blk(x, attn_mask)
        if self.downsample is not None:
            x_down = self.downsample(x, H, W)
            Wh, Ww = (H + 1) // 2, (W + 1) // 2
            return x, H, W, x_down, Wh, Ww
        else:
            return x, H, W, x, H, W

其中，block是对Swin Transformer Block的封装：

self.blocks = nn.ModuleList([
            SwinTransformerBlock(
                dim=dim,
                num_heads=num_heads,
                window_size=window_size,
                shift_size=0 if (i % 2 == 0) else window_size // 2,
                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,
                norm_layer=norm_layer)

此外，在正式进入Swin Transformer Block之前，还有一些其他的操作，如：

mask_windows = window_partition(img_mask, self.window_size)

该模块的作用是添加掩码，具体来看：

window_partition

window partition函数是用于对张量划分窗口，指定窗口大小。将原本的张量从 N H W C, 划分成 num_windows*B, window_size, window_size, C，其中 num_windows = H*W / window_size*window_size，即窗口的个数。而window reverse函数则是对应的逆过程。这两个函数会在后面的Window Attention用到。

在这里插入图片描述

def window_partition(x, window_size):
    """
    Args:
        x: (B, H, W, C)
        window_size (int): window size
    Returns:
        windows: (num_windows*B, window_size, window_size, C)
    """
    B, H, W, C = x.shape
    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)
    windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C)
    return windows

在这里插入图片描述

def window_reverse(windows, window_size, H, W):
    """
    Args:
        windows: (num_windows*B, window_size, window_size, C)
        window_size (int): Window size
        H (int): Height of image
        W (int): Width of image
    Returns:
        x: (B, H, W, C)
    """
    B = int(windows.shape[0] / (H * W / window_size / window_size))
    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)
    x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)
    return x

接下来我们便要正式进入到SwinTransformer Block模块了

SwinTransformer Block

在Swin Transformer Block中，最重要的莫过于Window Attention（W-MSA)与 Shifted Window Attention（SW-MSA)
传统的 Transformer 都是基于全局来计算注意力的，因此计算复杂度十分高。而 Swin Transformer 则将注意力的计算限制在每个窗口内，进而减少了计算量。其计算公式如下：
在这里插入图片描述
主要区别是在原始计算 Attention 的公式中的 Q,K 时加入了相对位置编码（+B)。

在这里插入图片描述

这里的维度变换较为复杂，使用DeBug来帮助理解，这里只运行第一阶段Swin Transformer Block

WindowAttention方法

def forward(self, x, mask=None):
        """ Forward function.
        Args:
            x: input features with shape of (num_windows*B, N, C)
            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None
        """
        B_, N, C = x.shape
        qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // 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)
        q = q * self.scale
        attn = (q @ k.transpose(-2, -1))
        relative_position_bias = self.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)  # Wh*Ww,Wh*Ww,nH
        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
        attn = attn + relative_position_bias.unsqueeze(0)
        if mask is not None:
            nW = mask.shape[0]
            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)
            attn = attn.view(-1, self.num_heads, N, N)
            attn = self.softmax(attn)
        else:
            attn = self.softmax(attn)
        attn = self.attn_drop(attn)
        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x

首先输入张量 x 形状为 [numWindows*B, window_size * window_size, C]

在这里插入图片描述

彭祥.

关注

21
点赞
踩
8

收藏

觉得还不错? 一键收藏
打赏
0
评论
SwinTransformer学习记录（二）之SwinTransformer Block

在前面的章节中，我们学习了SwinTransformer的整体框架，其主要由Patch Merging模块与SwinTansformer Block模块组成，
复制链接

扫一扫