Vision Transformer_mlp head是什么

作者：神奇cpp | 2024-07-24 14:31:47

踩

mlp head是什么

Vision Transformer模型详解

模型由三个模块组成

Linear Projection of Flattened Patches(Embeddind层)

Transformer Encoder(编码层)

MLP Head(最终用于分类的层结构)

Embedding层结构详解

对于标准的Transformer模块，要求输入的是token(向量)序列，即二维矩阵[num_token, token_dim]，如下图，token0-9对应的都是向量，以ViT-B/16为例，每个token向量长度为768.

对于图像数据而言，其数据格式为[H,W,C]是三维矩阵明显不是Transformer想要的。所以需要先通过一个Embedding层来对数据做个变换。如下图所示，首先将一个图片按给定大小分成一堆Patches。以ViT-B/16为例，将输入图片(224*224)按照16*16大小的patch进行划分，划分后会得到（224/16）^2=196个patches。接着通过线性映射将每个Patch映射到一维向量中，以ViT-B/16为例，每个Patches数据shape为[16,16,3]通过映射得到一个长度为768的向量(后面都直接称为Token)。[16,16,3]->[768]

在代码实现中，直接通过一个卷积层来实现。以ViT-B/16为例，直接使用一个卷积核大小为16*16，步距为16，卷积核个数为768的卷积实现。通过卷积[224,224,3]->[14,14,768]，然后把H以及W两个维度展平即可[14,14,768]->[196,768]，此时正好变成一个二维矩阵，正是Transformer想要的。

在输入Transformer Encoder之前注意需要加上[class]token以及Position Embedding。在原论文中，作者说参考BERT，在刚刚得到的一堆tokens中插入一个专门用于分类的[class]token，这个[class]token是一个可训练的参数，数据格式和其他token一样都是一个向量，以ViT-B/16为例，就是一个长度为768的向量，与之前从图片中生成的tokens拼接在一起，Cat([1,768],[196,768])->[197,768]。然后关于Position Embedding就是之前Transformer中讲到的Positional Encoding，这里的Position Embedding采用的是一个可训练的参数1D Pos. Emb.，是直接叠加在tokens上的(add)，所以shape要一样。以ViT-B/16为例，刚刚拼接[class]token后shape是[197,768],那么这里的Position Embedding的shape也是[197,768]。

对于Position Embedding作者也有一系列对比实验，在源码中默认使用的是1D Pos. Emb.,对比不使用Position Embedding准确率提升了大概3个点，和2D Pos. Emb.比起来没太大差别。

Transformer Encoder详解

其实就是重复堆叠Encoder Block L次，主要由以下几部分组成：

Layer Norm，这种Normalization方法主要针对NLP领域提出的，这里是对每个token进行Norm处理，之前也有讲过Layer Norm。

Multi-Head Attention

Dropout/DropPath，在原论文的代码是直接使用的Dropout层

MLP Block，如图右侧，就是全连接+GELU激活函数+Dropout组成，需要注意的是第一个全连接层会把输入节点的个数翻4倍[197,768]->[197,3072]，第二个全连接层会还原回原节点个数[197,3072]->[197,768]

MLP Head详解

上面通过Transformer Encoder后输出的shape和输入的shape是保持不变的，以ViT-B/16为例，输入的是[197,768]，输出的还是[197,768]。注意，在Transformer Encoder后其实还有一个Layer Norm没有画出来，后面画由ViT模型的详细结构。这里只是需要分类的信息，所以我们只需要提取出[class]token生成的对应结果就行，即[197,768]中抽取出[class]token对应的[1,768]。

接着我们通过MLP Head得到我们最终的分类结果。MLP Head原论文中说在训练ImageNet21K时是由Linear+tanh激活函数+Linear组成。但是迁移到ImageNet1K或者自己的数据集上时，只用一个Linear即可。

ViT Transformer网络结构

Hybrid模型详解

将传统CNN特征提取和Transformer进行结合。下面绘制的是以ResNet50作为特征提取器的混合模型，但这里的Resnet与之前的Resnet不同。首先这里的R50的卷积采用的StdConv2d不是传统的Conv2d,然后将所有的BatchNorm层换成了GroupNorm层。在原Resnet50网络中，stage1重复堆叠3次，stage2重复堆叠4次，stage3重复堆叠6次，stage4重复堆叠3次，但在这里的R50中，把stage4中的3个Block移至stage3中，所以stage3共重复堆叠9次。

通过R50 Backbone进行特征提取后，得到的特征矩阵shape是[14,14,1024]，接着再输入patch Embedding层中，注意Patch Embedding中卷积层Conv2d的kernel_size和stride都变成了1，只是用来调整channel。后面的部分和前面的ViT中讲的完全一样。

下表是论文用来对比ViT,Resnet（和刚刚讲的一样，使用的卷积层和Norm层都进行了修改）以及Hybrid模型的效果。通过对比发现，在训练epoch较少时Hybrid优于ViT，但当epoch增大后ViT优于Hybrid。

ViT模型的搭建参数

在论文Table1中给出是哪个模型（base/large/huge）的参数，在源码中除了有Patch Size为16*16的还有32*32的。其中的Layers就是Transformer Encoder中重复堆叠Encoder Block的次数，Hidden Size就是对应通过Embedding层后每个token的dim（向量的长度），MLP Size是Transformer Encoder中MLP Block第一个全连接的节点个数（是Hidden Size的四倍），Heads代表Transformer中Multi-Head Attention的heads数。

Model	Patch Size	Layers	Hidden Size D	MLP size	Heads	Params
ViT-Base	16x16	12	768	3072	12	86M
ViT-Large	16x16	24	1024	4096	16	307M
ViT-Huge	14x14	32	1280	5120	16	632M

ViT模型


"""
original code from rwightman:
https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
"""
from functools import partial
from collections import OrderedDict
 
import torch
import torch.nn as nn
 
 
# 随机深度
def drop_path(x, drop_prob: float = 0., training: bool = False):
    """
    Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
    This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
    the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
    See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for
    changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use
    'survival rate' as the argument.
    """
    if drop_prob == 0. or not training:
        return x
    keep_prob = 1 - drop_prob
    shape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
    random_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device)
    random_tensor.floor_()  # binarize
    output = x.div(keep_prob) * random_tensor
    return output
 
 
class DropPath(nn.Module):
    """
    Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks).
    """
    def __init__(self, drop_prob=None):
        super(DropPath, self).__init__()
        self.drop_prob = drop_prob
 
    def forward(self, x):
        return drop_path(x, self.drop_prob, self.training)
 
 
class PatchEmbed(nn.Module):
    """
    2D Image to Patch Embedding
    """
    # 将输入图像转化为self-attention的token向量格式
    # 输入图片224*224*3(RGB 3通道)，按照16*16的patch划分
    # 224,224,3-->14,14,768
    def __init__(self, img_size=224, patch_size=16, in_c=3, embed_dim=768, norm_layer=None):
        super().__init__()
        img_size = (img_size, img_size)
        patch_size = (patch_size, patch_size)
        self.img_size = img_size
        self.patch_size = patch_size
        # 224 // 16 ，224 // 16
        self.grid_size = (img_size[0] // patch_size[0], img_size[1] // patch_size[1])
        # 计算patches的数目 14 * 14
        self.num_patches = self.grid_size[0] * self.grid_size[1]
 
        # 卷积层 conv16*16,s=16,o = [(i+2p-k)/s] + 1=[(256-16)/16]+1=14,14*14
        self.proj = nn.Conv2d(in_c, embed_dim, kernel_size=patch_size, stride=patch_size)
        # nn.Identity()不做任何操作
        self.norm = norm_layer(embed_dim) if norm_layer else nn.Identity()
 
    # 正向传播过程：传入图片数据 VIT模型中图像输入大小是固定的，无法更改
    def forward(self, x):
        B, C, H, W = x.shape
        # 检查传入图片大小是否符合预先设定
        assert H == self.img_size[0] and W == self.img_size[1], \
            f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
 
        # flatten: [B, C, H, W] -> [B, C, HW] [B, 768, 14*14]
        # transpose: [B, C, HW] -> [B, HW, C] [B, 196, 768]
        x = self.proj(x).flatten(2).transpose(1, 2)
        x = self.norm(x)
        return x
 
 
# multihead -self attention
class Attention(nn.Module):
    def __init__(self,
                 dim,   # 输入token的dim = 768
                 num_heads=8,
                 qkv_bias=False,
                 qk_scale=None, # 根号下d
                 attn_drop_ratio=0.,
                 proj_drop_ratio=0.):
        super(Attention, self).__init__()
        self.num_heads = num_heads
        # 计算每个head qkv分得的头个数
        head_dim = dim // num_heads
        # 对应根号下d
        self.scale = qk_scale or head_dim ** -0.5
        # 全连接层实现qkv，严格来说是三个分开，这里提高并行化
        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop_ratio)
        self.proj = nn.Linear(dim, dim)         # 要将计算完的分头全连接起来，wo映射通过linear实现
        self.proj_drop = nn.Dropout(proj_drop_ratio)
 
    # 经典
    def forward(self, x):
        # [batch_size, num_patches + 1, total_embed_dim]
        B, N, C = x.shape
 
        # qkv(): -> [batch_size, num_patches + 1, 3 * total_embed_dim]
        # reshape进行拆分: -> [batch_size, num_patches + 1, 3(代表qkv), num_heads, embed_dim_per_head]
        # permute: -> [3, batch_size, num_heads, num_patches + 1, embed_dim_per_head] 方便运算
        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
        # [batch_size, num_heads, num_patches + 1, embed_dim_per_head]
        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)
 
        # transpose: -> [batch_size, num_heads, embed_dim_per_head, num_patches + 1]
        # @：矩阵乘法
        # @: multiply -> [batch_size, num_heads, num_patches + 1, num_patches + 1]
        attn = (q @ k.transpose(-2, -1)) * self.scale # scale进行norm处理
        # dim=-1 针对每一行进行处理
        attn = attn.softmax(dim=-1)
        attn = self.attn_drop(attn)
 
        # @: multiply -> [batch_size, num_heads, num_patches + 1, embed_dim_per_head]
        # transpose: -> [batch_size, num_patches + 1, num_heads, embed_dim_per_head]
        # reshape: -> [batch_size, num_patches + 1, total_embed_dim]  把最后两个信息拼接在一起
        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x
 
 
class Mlp(nn.Module):
    """
    MLP as used in Vision Transformer, MLP-Mixer and related networks
    """
    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) # 增加通道数
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features) # 还原通道数
        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
 
 
# transformer中就是将block重复堆叠L次
class Block(nn.Module):
    def __init__(self,
                 dim,
                 num_heads,
                 mlp_ratio=4., # 第一个全连接层是输入节点个数的4倍
                 qkv_bias=False,
                 qk_scale=None,
                 drop_ratio=0.,
                 attn_drop_ratio=0.,
                 drop_path_ratio=0.,
                 act_layer=nn.GELU,
                 norm_layer=nn.LayerNorm):
        super(Block, self).__init__()
        self.norm1 = norm_layer(dim)
        self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale,
                              attn_drop_ratio=attn_drop_ratio, proj_drop_ratio=drop_ratio)
        # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
        self.drop_path = DropPath(drop_path_ratio) if drop_path_ratio > 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_ratio)
 
    def forward(self, x):
        x = x + self.drop_path(self.attn(self.norm1(x)))
        x = x + self.drop_path(self.mlp(self.norm2(x)))
        return x
 
 
class VisionTransformer(nn.Module):
    def __init__(self, img_size=224, patch_size=16, in_c=3, num_classes=1000,
                 embed_dim=768, depth=12, num_heads=12, mlp_ratio=4.0, qkv_bias=True,
                 qk_scale=None, representation_size=None, distilled=False, drop_ratio=0.,
                 attn_drop_ratio=0., drop_path_ratio=0., embed_layer=PatchEmbed, norm_layer=None,
                 act_layer=None):
        """
        Args:
            img_size (int, tuple): input image size
            patch_size (int, tuple): patch size
            in_c (int): number of input channels
            num_classes (int): number of classes for classification head
            embed_dim (int): embedding dimension
            depth (int): depth of transformer 重复堆叠encoder的次数
            num_heads (int): number of attention heads
            mlp_ratio (int): ratio of mlp hidden dim to embedding dim
            qkv_bias (bool): enable bias for qkv if True
            qk_scale (float): override default qk scale of head_dim ** -0.5 if set
            representation_size (Optional[int]): enable and set representation layer (pre-logits) to this value if set
            distilled (bool): model includes a distillation token and head as in DeiT models
            drop_ratio (float): dropout rate
            attn_drop_ratio (float): attention dropout rate
            drop_path_ratio (float): stochastic depth rate
            embed_layer (nn.Module): patch embedding layer
            norm_layer: (nn.Module): normalization layer
        """
        super(VisionTransformer, self).__init__()
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
        self.num_tokens = 2 if distilled else 1
        norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
        act_layer = act_layer or nn.GELU
 
        self.patch_embed = embed_layer(img_size=img_size, patch_size=patch_size, in_c=in_c, embed_dim=embed_dim)
        num_patches = self.patch_embed.num_patches
 
        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
        self.dist_token = nn.Parameter(torch.zeros(1, 1, embed_dim)) if distilled else None
        self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + self.num_tokens, embed_dim))
        self.pos_drop = nn.Dropout(p=drop_ratio)
 
        dpr = [x.item() for x in torch.linspace(0, drop_path_ratio, depth)]  # stochastic depth decay rule
        self.blocks = nn.Sequential(*[
            Block(dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                  drop_ratio=drop_ratio, attn_drop_ratio=attn_drop_ratio, drop_path_ratio=dpr[i],
                  norm_layer=norm_layer, act_layer=act_layer)
            for i in range(depth)
        ])
        self.norm = norm_layer(embed_dim)
 
        # Representation layer
        if representation_size and not distilled:
            self.has_logits = True
            self.num_features = representation_size
            self.pre_logits = nn.Sequential(OrderedDict([
                ("fc", nn.Linear(embed_dim, representation_size)),
                ("act", nn.Tanh())
            ]))
        else:
            self.has_logits = False
            self.pre_logits = nn.Identity()
 
        # Classifier head(s)
        self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()
        self.head_dist = None
        if distilled:
            self.head_dist = nn.Linear(self.embed_dim, self.num_classes) if num_classes > 0 else nn.Identity()
 
        # Weight init
        nn.init.trunc_normal_(self.pos_embed, std=0.02)
        if self.dist_token is not None:
            nn.init.trunc_normal_(self.dist_token, std=0.02)
 
        nn.init.trunc_normal_(self.cls_token, std=0.02)
        self.apply(_init_vit_weights)
 
    # 正向传播过程
    def forward_features(self, x):
        # [B, C, H, W] -> [B, num_patches, embed_dim]
        x = self.patch_embed(x)  # [B, 196, 768]
        # [1, 1, 768] -> [B, 1, 768]
        cls_token = self.cls_token.expand(x.shape[0], -1, -1)
        if self.dist_token is None:
            x = torch.cat((cls_token, x), dim=1)  # [B, 197, 768]
        else:
            x = torch.cat((cls_token, self.dist_token.expand(x.shape[0], -1, -1), x), dim=1)
 
        x = self.pos_drop(x + self.pos_embed)
        x = self.blocks(x)
        x = self.norm(x)
        if self.dist_token is None:
            return self.pre_logits(x[:, 0])
        else:
            return x[:, 0], x[:, 1]
 
    def forward(self, x):
        x = self.forward_features(x)
        if self.head_dist is not None:
            x, x_dist = self.head(x[0]), self.head_dist(x[1])
            if self.training and not torch.jit.is_scripting():
                # during inference, return the average of both classifier predictions
                return x, x_dist
            else:
                return (x + x_dist) / 2
        else:
            x = self.head(x)
        return x
 
 
def _init_vit_weights(m):
    """
    ViT weight initialization
    :param m: module
    """
    if isinstance(m, nn.Linear):
        nn.init.trunc_normal_(m.weight, std=.01)
        if m.bias is not None:
            nn.init.zeros_(m.bias)
    elif isinstance(m, nn.Conv2d):
        nn.init.kaiming_normal_(m.weight, mode="fan_out")
        if m.bias is not None:
            nn.init.zeros_(m.bias)
    elif isinstance(m, nn.LayerNorm):
        nn.init.zeros_(m.bias)
        nn.init.ones_(m.weight)
 
 
def vit_base_patch16_224(num_classes: int = 1000):
    """
    ViT-Base model (ViT-B/16) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-1k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    链接: https://pan.baidu.com/s/1zqb08naP0RPqqfSXfkB2EA  密码: eu9f
    """
    model = VisionTransformer(img_size=224,
                              patch_size=16,
                              embed_dim=768,
                              depth=12,
                              num_heads=12,
                              representation_size=None,
                              num_classes=num_classes)
    return model
 
 
def vit_base_patch16_224_in21k(num_classes: int = 21843, has_logits: bool = True):
    """
    ViT-Base model (ViT-B/16) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_patch16_224_in21k-e5005f0a.pth
    """
    model = VisionTransformer(img_size=224,
                              patch_size=16,
                              embed_dim=768,
                              depth=12,
                              num_heads=12,
                              representation_size=768 if has_logits else None,
                              num_classes=num_classes)
    return model
 
 
def vit_base_patch32_224(num_classes: int = 1000):
    """
    ViT-Base model (ViT-B/32) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-1k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    链接: https://pan.baidu.com/s/1hCv0U8pQomwAtHBYc4hmZg  密码: s5hl
    """
    model = VisionTransformer(img_size=224,
                              patch_size=32,
                              embed_dim=768,
                              depth=12,
                              num_heads=12,
                              representation_size=None,
                              num_classes=num_classes)
    return model
 
 
def vit_base_patch32_224_in21k(num_classes: int = 21843, has_logits: bool = True):
    """
    ViT-Base model (ViT-B/32) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_patch32_224_in21k-8db57226.pth
    """
    model = VisionTransformer(img_size=224,
                              patch_size=32,
                              embed_dim=768,
                              depth=12,
                              num_heads=12,
                              representation_size=768 if has_logits else None,
                              num_classes=num_classes)
    return model
 
 
def vit_large_patch16_224(num_classes: int = 1000):
    """
    ViT-Large model (ViT-L/16) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-1k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    链接: https://pan.baidu.com/s/1cxBgZJJ6qUWPSBNcE4TdRQ  密码: qqt8
    """
    model = VisionTransformer(img_size=224,
                              patch_size=16,
                              embed_dim=1024,
                              depth=24,
                              num_heads=16,
                              representation_size=None,
                              num_classes=num_classes)
    return model
 
 
def vit_large_patch16_224_in21k(num_classes: int = 21843, has_logits: bool = True):
    """
    ViT-Large model (ViT-L/16) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_patch16_224_in21k-606da67d.pth
    """
    model = VisionTransformer(img_size=224,
                              patch_size=16,
                              embed_dim=1024,
                              depth=24,
                              num_heads=16,
                              representation_size=1024 if has_logits else None,
                              num_classes=num_classes)
    return model
 
 
def vit_large_patch32_224_in21k(num_classes: int = 21843, has_logits: bool = True):
    """
    ViT-Large model (ViT-L/32) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    weights ported from official Google JAX impl:
    https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_patch32_224_in21k-9046d2e7.pth
    """
    model = VisionTransformer(img_size=224,
                              patch_size=32,
                              embed_dim=1024,
                              depth=24,
                              num_heads=16,
                              representation_size=1024 if has_logits else None,
                              num_classes=num_classes)
    return model
 
 
def vit_huge_patch14_224_in21k(num_classes: int = 21843, has_logits: bool = True):
    """
    ViT-Huge model (ViT-H/14) from original paper (https://arxiv.org/abs/2010.11929).
    ImageNet-21k weights @ 224x224, source https://github.com/google-research/vision_transformer.
    NOTE: converted weights not currently available, too large for github release hosting.
    """
    model = VisionTransformer(img_size=224,
                              patch_size=14,
                              embed_dim=1280,
                              depth=32,
                              num_heads=16,
                              representation_size=1280 if has_logits else None,
                              num_classes=num_classes)
    return model

本文内容由网友自发贡献，转载请注明出处：【wpsshop博客】