fast-reid/fastreid/modeling/backbones/vision_transformer.py

400 lines
16 KiB
Python

""" Vision Transformer (ViT) in PyTorch
A PyTorch implement of Vision Transformers as described in
'An Image Is Worth 16 x 16 Words: Transformers for Image Recognition at Scale' - https://arxiv.org/abs/2010.11929
The official jax code is released and available at https://github.com/google-research/vision_transformer
Status/TODO:
* Models updated to be compatible with official impl. Args added to support backward compat for old PyTorch weights.
* Weights ported from official jax impl for 384x384 base and small models, 16x16 and 32x32 patches.
* Trained (supervised on ImageNet-1k) my custom 'small' patch model to 77.9, 'base' to 79.4 top-1 with this code.
* Hopefully find time and GPUs for SSL or unsupervised pretraining on OpenImages w/ ImageNet fine-tune in future.
Acknowledgments:
* The paper authors for releasing code and weights, thanks!
* I fixed my class token impl based on Phil Wang's https://github.com/lucidrains/vit-pytorch ... check it out
for some einops/einsum fun
* Simple transformer style inspired by Andrej Karpathy's https://github.com/karpathy/minGPT
* Bert reference code checks against Huggingface Transformers and Tensorflow Bert
Hacked together by / Copyright 2020 Ross Wightman
"""
import logging
import math
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from fastreid.layers import DropPath, trunc_normal_, to_2tuple
from fastreid.utils.checkpoint import get_missing_parameters_message, get_unexpected_parameters_message
from .build import BACKBONE_REGISTRY
logger = logging.getLogger(__name__)
class Mlp(nn.Module):
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
# NOTE scale factor was wrong in my original version, can set manually to be compat with prev weights
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):
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)
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(B, N, C)
x = self.proj(x)
x = self.proj_drop(x)
return x
class Block(nn.Module):
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):
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)
# NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
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)
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 PatchEmbed(nn.Module):
""" Image to Patch Embedding
"""
def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
super().__init__()
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0])
self.img_size = img_size
self.patch_size = patch_size
self.num_patches = num_patches
self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
def forward(self, x):
B, C, H, W = x.shape
# FIXME look at relaxing size constraints
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]})."
x = self.proj(x).flatten(2).transpose(1, 2)
return x
class HybridEmbed(nn.Module):
""" CNN Feature Map Embedding
Extract feature map from CNN, flatten, project to embedding dim.
"""
def __init__(self, backbone, img_size=224, feature_size=None, in_chans=3, embed_dim=768):
super().__init__()
assert isinstance(backbone, nn.Module)
img_size = to_2tuple(img_size)
self.img_size = img_size
self.backbone = backbone
if feature_size is None:
with torch.no_grad():
# FIXME this is hacky, but most reliable way of determining the exact dim of the output feature
# map for all networks, the feature metadata has reliable channel and stride info, but using
# stride to calc feature dim requires info about padding of each stage that isn't captured.
training = backbone.training
if training:
backbone.eval()
o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))
if isinstance(o, (list, tuple)):
o = o[-1] # last feature if backbone outputs list/tuple of features
feature_size = o.shape[-2:]
feature_dim = o.shape[1]
backbone.train(training)
else:
feature_size = to_2tuple(feature_size)
if hasattr(self.backbone, 'feature_info'):
feature_dim = self.backbone.feature_info.channels()[-1]
else:
feature_dim = self.backbone.num_features
self.num_patches = feature_size[0] * feature_size[1]
self.proj = nn.Conv2d(feature_dim, embed_dim, 1)
def forward(self, x):
x = self.backbone(x)
if isinstance(x, (list, tuple)):
x = x[-1] # last feature if backbone outputs list/tuple of features
x = self.proj(x).flatten(2).transpose(1, 2)
return x
class PatchEmbed_overlap(nn.Module):
""" Image to Patch Embedding with overlapping patches
"""
def __init__(self, img_size=224, patch_size=16, stride_size=20, in_chans=3, embed_dim=768):
super().__init__()
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
stride_size_tuple = to_2tuple(stride_size)
self.num_x = (img_size[1] - patch_size[1]) // stride_size_tuple[1] + 1
self.num_y = (img_size[0] - patch_size[0]) // stride_size_tuple[0] + 1
num_patches = self.num_x * self.num_y
self.img_size = img_size
self.patch_size = patch_size
self.num_patches = num_patches
self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=stride_size)
for m in self.modules():
if isinstance(m, nn.Conv2d):
n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
m.weight.data.normal_(0, math.sqrt(2. / n))
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
elif isinstance(m, nn.InstanceNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def forward(self, x):
B, C, H, W = x.shape
# FIXME look at relaxing size constraints
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]})."
x = self.proj(x)
x = x.flatten(2).transpose(1, 2) # [64, 8, 768]
return x
class VisionTransformer(nn.Module):
""" Vision Transformer
A PyTorch impl of : `An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale`
- https://arxiv.org/abs/2010.11929
Includes distillation token & head support for `DeiT: Data-efficient Image Transformers`
- https://arxiv.org/abs/2012.12877
"""
def __init__(self, img_size=224, patch_size=16, stride_size=16, in_chans=3, embed_dim=768,
depth=12, num_heads=12, mlp_ratio=4., qkv_bias=False, qk_scale=None,
drop_rate=0., attn_drop_rate=0., camera=0, drop_path_rate=0., hybrid_backbone=None,
norm_layer=partial(nn.LayerNorm, eps=1e-6), sie_xishu=1.0):
super().__init__()
self.num_features = self.embed_dim = embed_dim # num_features for consistency with other models
if hybrid_backbone is not None:
self.patch_embed = HybridEmbed(
hybrid_backbone, img_size=img_size, in_chans=in_chans, embed_dim=embed_dim)
else:
self.patch_embed = PatchEmbed_overlap(
img_size=img_size, patch_size=patch_size, stride_size=stride_size, in_chans=in_chans,
embed_dim=embed_dim)
num_patches = self.patch_embed.num_patches
self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
self.cam_num = camera
self.sie_xishu = sie_xishu
# Initialize SIE Embedding
if camera > 1:
self.sie_embed = nn.Parameter(torch.zeros(camera, 1, embed_dim))
trunc_normal_(self.sie_embed, std=.02)
self.pos_drop = nn.Dropout(p=drop_rate)
dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)] # stochastic depth decay rule
self.blocks = nn.ModuleList([
Block(
dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer)
for i in range(depth)])
self.norm = norm_layer(embed_dim)
trunc_normal_(self.cls_token, std=.02)
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)
@torch.jit.ignore
def no_weight_decay(self):
return {'pos_embed', 'cls_token'}
def forward(self, x, camera_id=None):
B = x.shape[0]
x = self.patch_embed(x)
cls_tokens = self.cls_token.expand(B, -1, -1) # stole cls_tokens impl from Phil Wang, thanks
x = torch.cat((cls_tokens, x), dim=1)
if self.cam_num > 0:
x = x + self.pos_embed + self.sie_xishu * self.sie_embed[camera_id]
else:
x = x + self.pos_embed
x = self.pos_drop(x)
for blk in self.blocks:
x = blk(x)
x = self.norm(x)
return x[:, 0].reshape(x.shape[0], -1, 1, 1)
def resize_pos_embed(posemb, posemb_new, hight, width):
# Rescale the grid of position embeddings when loading from state_dict. Adapted from
# https://github.com/google-research/vision_transformer/blob/00883dd691c63a6830751563748663526e811cee/vit_jax/checkpoint.py#L224
ntok_new = posemb_new.shape[1]
posemb_token, posemb_grid = posemb[:, :1], posemb[0, 1:]
ntok_new -= 1
gs_old = int(math.sqrt(len(posemb_grid)))
logger.info('Resized position embedding from size:{} to size: {} with height:{} width: {}'.format(posemb.shape,
posemb_new.shape,
hight,
width))
posemb_grid = posemb_grid.reshape(1, gs_old, gs_old, -1).permute(0, 3, 1, 2)
posemb_grid = F.interpolate(posemb_grid, size=(hight, width), mode='bilinear')
posemb_grid = posemb_grid.permute(0, 2, 3, 1).reshape(1, hight * width, -1)
posemb = torch.cat([posemb_token, posemb_grid], dim=1)
return posemb
@BACKBONE_REGISTRY.register()
def build_vit_backbone(cfg):
"""
Create a Vision Transformer instance from config.
Returns:
SwinTransformer: a :class:`SwinTransformer` instance.
"""
# fmt: off
input_size = cfg.INPUT.SIZE_TRAIN
pretrain = cfg.MODEL.BACKBONE.PRETRAIN
pretrain_path = cfg.MODEL.BACKBONE.PRETRAIN_PATH
depth = cfg.MODEL.BACKBONE.DEPTH
sie_xishu = cfg.MODEL.BACKBONE.SIE_COE
stride_size = cfg.MODEL.BACKBONE.STRIDE_SIZE
drop_ratio = cfg.MODEL.BACKBONE.DROP_RATIO
drop_path_ratio = cfg.MODEL.BACKBONE.DROP_PATH_RATIO
attn_drop_rate = cfg.MODEL.BACKBONE.ATT_DROP_RATE
# fmt: on
num_depth = {
'small': 8,
'base': 12,
}[depth]
num_heads = {
'small': 8,
'base': 12,
}[depth]
mlp_ratio = {
'small': 3.,
'base': 4.
}[depth]
qkv_bias = {
'small': False,
'base': True
}[depth]
qk_scale = {
'small': 768 ** -0.5,
'base': None,
}[depth]
model = VisionTransformer(img_size=input_size, sie_xishu=sie_xishu, stride_size=stride_size, depth=num_depth,
num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
drop_path_rate=drop_path_ratio, drop_rate=drop_ratio, attn_drop_rate=attn_drop_rate)
if pretrain:
try:
state_dict = torch.load(pretrain_path, map_location=torch.device('cpu'))
logger.info(f"Loading pretrained model from {pretrain_path}")
if 'model' in state_dict:
state_dict = state_dict.pop('model')
if 'state_dict' in state_dict:
state_dict = state_dict.pop('state_dict')
for k, v in state_dict.items():
if 'head' in k or 'dist' in k:
continue
if 'patch_embed.proj.weight' in k and len(v.shape) < 4:
# For old models that I trained prior to conv based patchification
O, I, H, W = model.patch_embed.proj.weight.shape
v = v.reshape(O, -1, H, W)
elif k == 'pos_embed' and v.shape != model.pos_embed.shape:
# To resize pos embedding when using model at different size from pretrained weights
if 'distilled' in pretrain_path:
logger.info("distill need to choose right cls token in the pth.")
v = torch.cat([v[:, 0:1], v[:, 2:]], dim=1)
v = resize_pos_embed(v, model.pos_embed.data, model.patch_embed.num_y, model.patch_embed.num_x)
state_dict[k] = v
except FileNotFoundError as e:
logger.info(f'{pretrain_path} is not found! Please check this path.')
raise e
except KeyError as e:
logger.info("State dict keys error! Please check the state dict.")
raise e
incompatible = model.load_state_dict(state_dict, strict=False)
if incompatible.missing_keys:
logger.info(
get_missing_parameters_message(incompatible.missing_keys)
)
if incompatible.unexpected_keys:
logger.info(
get_unexpected_parameters_message(incompatible.unexpected_keys)
)
return model