ZhengPeng7
commited on
Commit
·
6643d46
1
Parent(s):
ab9a192
Add the deployment option. Update the model codes.
Browse files- birefnet.py +29 -26
- handler.py +9 -4
birefnet.py
CHANGED
@@ -615,6 +615,7 @@ from timm.models.layers import DropPath, to_2tuple, trunc_normal_
|
|
615 |
|
616 |
# config = Config()
|
617 |
|
|
|
618 |
class Mlp(nn.Module):
|
619 |
""" Multilayer perceptron."""
|
620 |
|
@@ -739,7 +740,8 @@ class WindowAttention(nn.Module):
|
|
739 |
attn = (q @ k.transpose(-2, -1))
|
740 |
|
741 |
relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
|
742 |
-
self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1
|
|
|
743 |
relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
744 |
attn = attn + relative_position_bias.unsqueeze(0)
|
745 |
|
@@ -974,8 +976,9 @@ class BasicLayer(nn.Module):
|
|
974 |
"""
|
975 |
|
976 |
# calculate attention mask for SW-MSA
|
977 |
-
|
978 |
-
|
|
|
979 |
img_mask = torch.zeros((1, Hp, Wp, 1), device=x.device) # 1 Hp Wp 1
|
980 |
h_slices = (slice(0, -self.window_size),
|
981 |
slice(-self.window_size, -self.shift_size),
|
@@ -1961,6 +1964,7 @@ import torch.nn as nn
|
|
1961 |
import torch.nn.functional as F
|
1962 |
from kornia.filters import laplacian
|
1963 |
from transformers import PreTrainedModel
|
|
|
1964 |
|
1965 |
# from config import Config
|
1966 |
# from dataset import class_labels_TR_sorted
|
@@ -1974,13 +1978,24 @@ from transformers import PreTrainedModel
|
|
1974 |
from .BiRefNet_config import BiRefNetConfig
|
1975 |
|
1976 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1977 |
class BiRefNet(
|
1978 |
PreTrainedModel
|
1979 |
):
|
1980 |
config_class = BiRefNetConfig
|
1981 |
def __init__(self, bb_pretrained=True, config=BiRefNetConfig()):
|
1982 |
-
super(BiRefNet, self).__init__(
|
1983 |
-
bb_pretrained = config.bb_pretrained
|
1984 |
self.config = Config()
|
1985 |
self.epoch = 1
|
1986 |
self.bb = build_backbone(self.config.bb, pretrained=bb_pretrained)
|
@@ -2124,18 +2139,6 @@ class Decoder(nn.Module):
|
|
2124 |
self.gdt_convs_attn_3 = nn.Sequential(nn.Conv2d(_N, 1, 1, 1, 0))
|
2125 |
self.gdt_convs_attn_2 = nn.Sequential(nn.Conv2d(_N, 1, 1, 1, 0))
|
2126 |
|
2127 |
-
def get_patches_batch(self, x, p):
|
2128 |
-
_size_h, _size_w = p.shape[2:]
|
2129 |
-
patches_batch = []
|
2130 |
-
for idx in range(x.shape[0]):
|
2131 |
-
columns_x = torch.split(x[idx], split_size_or_sections=_size_w, dim=-1)
|
2132 |
-
patches_x = []
|
2133 |
-
for column_x in columns_x:
|
2134 |
-
patches_x += [p.unsqueeze(0) for p in torch.split(column_x, split_size_or_sections=_size_h, dim=-2)]
|
2135 |
-
patch_sample = torch.cat(patches_x, dim=1)
|
2136 |
-
patches_batch.append(patch_sample)
|
2137 |
-
return torch.cat(patches_batch, dim=0)
|
2138 |
-
|
2139 |
def forward(self, features):
|
2140 |
if self.training and self.config.out_ref:
|
2141 |
outs_gdt_pred = []
|
@@ -2146,10 +2149,10 @@ class Decoder(nn.Module):
|
|
2146 |
outs = []
|
2147 |
|
2148 |
if self.config.dec_ipt:
|
2149 |
-
patches_batch =
|
2150 |
x4 = torch.cat((x4, self.ipt_blk5(F.interpolate(patches_batch, size=x4.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2151 |
p4 = self.decoder_block4(x4)
|
2152 |
-
m4 = self.conv_ms_spvn_4(p4) if self.config.ms_supervision else None
|
2153 |
if self.config.out_ref:
|
2154 |
p4_gdt = self.gdt_convs_4(p4)
|
2155 |
if self.training:
|
@@ -2167,10 +2170,10 @@ class Decoder(nn.Module):
|
|
2167 |
_p3 = _p4 + self.lateral_block4(x3)
|
2168 |
|
2169 |
if self.config.dec_ipt:
|
2170 |
-
patches_batch =
|
2171 |
_p3 = torch.cat((_p3, self.ipt_blk4(F.interpolate(patches_batch, size=x3.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2172 |
p3 = self.decoder_block3(_p3)
|
2173 |
-
m3 = self.conv_ms_spvn_3(p3) if self.config.ms_supervision else None
|
2174 |
if self.config.out_ref:
|
2175 |
p3_gdt = self.gdt_convs_3(p3)
|
2176 |
if self.training:
|
@@ -2193,10 +2196,10 @@ class Decoder(nn.Module):
|
|
2193 |
_p2 = _p3 + self.lateral_block3(x2)
|
2194 |
|
2195 |
if self.config.dec_ipt:
|
2196 |
-
patches_batch =
|
2197 |
_p2 = torch.cat((_p2, self.ipt_blk3(F.interpolate(patches_batch, size=x2.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2198 |
p2 = self.decoder_block2(_p2)
|
2199 |
-
m2 = self.conv_ms_spvn_2(p2) if self.config.ms_supervision else None
|
2200 |
if self.config.out_ref:
|
2201 |
p2_gdt = self.gdt_convs_2(p2)
|
2202 |
if self.training:
|
@@ -2214,17 +2217,17 @@ class Decoder(nn.Module):
|
|
2214 |
_p1 = _p2 + self.lateral_block2(x1)
|
2215 |
|
2216 |
if self.config.dec_ipt:
|
2217 |
-
patches_batch =
|
2218 |
_p1 = torch.cat((_p1, self.ipt_blk2(F.interpolate(patches_batch, size=x1.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2219 |
_p1 = self.decoder_block1(_p1)
|
2220 |
_p1 = F.interpolate(_p1, size=x.shape[2:], mode='bilinear', align_corners=True)
|
2221 |
|
2222 |
if self.config.dec_ipt:
|
2223 |
-
patches_batch =
|
2224 |
_p1 = torch.cat((_p1, self.ipt_blk1(F.interpolate(patches_batch, size=x.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2225 |
p1_out = self.conv_out1(_p1)
|
2226 |
|
2227 |
-
if self.config.ms_supervision:
|
2228 |
outs.append(m4)
|
2229 |
outs.append(m3)
|
2230 |
outs.append(m2)
|
|
|
615 |
|
616 |
# config = Config()
|
617 |
|
618 |
+
|
619 |
class Mlp(nn.Module):
|
620 |
""" Multilayer perceptron."""
|
621 |
|
|
|
740 |
attn = (q @ k.transpose(-2, -1))
|
741 |
|
742 |
relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
|
743 |
+
self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1
|
744 |
+
) # Wh*Ww, Wh*Ww, nH
|
745 |
relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
746 |
attn = attn + relative_position_bias.unsqueeze(0)
|
747 |
|
|
|
976 |
"""
|
977 |
|
978 |
# calculate attention mask for SW-MSA
|
979 |
+
# Turn int to torch.tensor for the compatiability with torch.compile in PyTorch 2.5.
|
980 |
+
Hp = torch.ceil(torch.tensor(H) / self.window_size).to(torch.int64) * self.window_size
|
981 |
+
Wp = torch.ceil(torch.tensor(W) / self.window_size).to(torch.int64) * self.window_size
|
982 |
img_mask = torch.zeros((1, Hp, Wp, 1), device=x.device) # 1 Hp Wp 1
|
983 |
h_slices = (slice(0, -self.window_size),
|
984 |
slice(-self.window_size, -self.shift_size),
|
|
|
1964 |
import torch.nn.functional as F
|
1965 |
from kornia.filters import laplacian
|
1966 |
from transformers import PreTrainedModel
|
1967 |
+
from einops import rearrange
|
1968 |
|
1969 |
# from config import Config
|
1970 |
# from dataset import class_labels_TR_sorted
|
|
|
1978 |
from .BiRefNet_config import BiRefNetConfig
|
1979 |
|
1980 |
|
1981 |
+
def image2patches(image, grid_h=2, grid_w=2, patch_ref=None, transformation='b c (hg h) (wg w) -> (b hg wg) c h w'):
|
1982 |
+
if patch_ref is not None:
|
1983 |
+
grid_h, grid_w = image.shape[-2] // patch_ref.shape[-2], image.shape[-1] // patch_ref.shape[-1]
|
1984 |
+
patches = rearrange(image, transformation, hg=grid_h, wg=grid_w)
|
1985 |
+
return patches
|
1986 |
+
|
1987 |
+
def patches2image(patches, grid_h=2, grid_w=2, patch_ref=None, transformation='(b hg wg) c h w -> b c (hg h) (wg w)'):
|
1988 |
+
if patch_ref is not None:
|
1989 |
+
grid_h, grid_w = patch_ref.shape[-2] // patches[0].shape[-2], patch_ref.shape[-1] // patches[0].shape[-1]
|
1990 |
+
image = rearrange(patches, transformation, hg=grid_h, wg=grid_w)
|
1991 |
+
return image
|
1992 |
+
|
1993 |
class BiRefNet(
|
1994 |
PreTrainedModel
|
1995 |
):
|
1996 |
config_class = BiRefNetConfig
|
1997 |
def __init__(self, bb_pretrained=True, config=BiRefNetConfig()):
|
1998 |
+
super(BiRefNet, self).__init__()
|
|
|
1999 |
self.config = Config()
|
2000 |
self.epoch = 1
|
2001 |
self.bb = build_backbone(self.config.bb, pretrained=bb_pretrained)
|
|
|
2139 |
self.gdt_convs_attn_3 = nn.Sequential(nn.Conv2d(_N, 1, 1, 1, 0))
|
2140 |
self.gdt_convs_attn_2 = nn.Sequential(nn.Conv2d(_N, 1, 1, 1, 0))
|
2141 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2142 |
def forward(self, features):
|
2143 |
if self.training and self.config.out_ref:
|
2144 |
outs_gdt_pred = []
|
|
|
2149 |
outs = []
|
2150 |
|
2151 |
if self.config.dec_ipt:
|
2152 |
+
patches_batch = image2patches(x, patch_ref=x4, transformation='b c (hg h) (wg w) -> b (c hg wg) h w') if self.split else x
|
2153 |
x4 = torch.cat((x4, self.ipt_blk5(F.interpolate(patches_batch, size=x4.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2154 |
p4 = self.decoder_block4(x4)
|
2155 |
+
m4 = self.conv_ms_spvn_4(p4) if self.config.ms_supervision and self.training else None
|
2156 |
if self.config.out_ref:
|
2157 |
p4_gdt = self.gdt_convs_4(p4)
|
2158 |
if self.training:
|
|
|
2170 |
_p3 = _p4 + self.lateral_block4(x3)
|
2171 |
|
2172 |
if self.config.dec_ipt:
|
2173 |
+
patches_batch = image2patches(x, patch_ref=_p3, transformation='b c (hg h) (wg w) -> b (c hg wg) h w') if self.split else x
|
2174 |
_p3 = torch.cat((_p3, self.ipt_blk4(F.interpolate(patches_batch, size=x3.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2175 |
p3 = self.decoder_block3(_p3)
|
2176 |
+
m3 = self.conv_ms_spvn_3(p3) if self.config.ms_supervision and self.training else None
|
2177 |
if self.config.out_ref:
|
2178 |
p3_gdt = self.gdt_convs_3(p3)
|
2179 |
if self.training:
|
|
|
2196 |
_p2 = _p3 + self.lateral_block3(x2)
|
2197 |
|
2198 |
if self.config.dec_ipt:
|
2199 |
+
patches_batch = image2patches(x, patch_ref=_p2, transformation='b c (hg h) (wg w) -> b (c hg wg) h w') if self.split else x
|
2200 |
_p2 = torch.cat((_p2, self.ipt_blk3(F.interpolate(patches_batch, size=x2.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2201 |
p2 = self.decoder_block2(_p2)
|
2202 |
+
m2 = self.conv_ms_spvn_2(p2) if self.config.ms_supervision and self.training else None
|
2203 |
if self.config.out_ref:
|
2204 |
p2_gdt = self.gdt_convs_2(p2)
|
2205 |
if self.training:
|
|
|
2217 |
_p1 = _p2 + self.lateral_block2(x1)
|
2218 |
|
2219 |
if self.config.dec_ipt:
|
2220 |
+
patches_batch = image2patches(x, patch_ref=_p1, transformation='b c (hg h) (wg w) -> b (c hg wg) h w') if self.split else x
|
2221 |
_p1 = torch.cat((_p1, self.ipt_blk2(F.interpolate(patches_batch, size=x1.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2222 |
_p1 = self.decoder_block1(_p1)
|
2223 |
_p1 = F.interpolate(_p1, size=x.shape[2:], mode='bilinear', align_corners=True)
|
2224 |
|
2225 |
if self.config.dec_ipt:
|
2226 |
+
patches_batch = image2patches(x, patch_ref=_p1, transformation='b c (hg h) (wg w) -> b (c hg wg) h w') if self.split else x
|
2227 |
_p1 = torch.cat((_p1, self.ipt_blk1(F.interpolate(patches_batch, size=x.shape[2:], mode='bilinear', align_corners=True))), 1)
|
2228 |
p1_out = self.conv_out1(_p1)
|
2229 |
|
2230 |
+
if self.config.ms_supervision and self.training:
|
2231 |
outs.append(m4)
|
2232 |
outs.append(m3)
|
2233 |
outs.append(m2)
|
handler.py
CHANGED
@@ -1,7 +1,11 @@
|
|
1 |
# These HF deployment codes refer to https://huggingface.co/not-lain/BiRefNet/raw/main/handler.py.
|
2 |
from typing import Dict, List, Any, Tuple
|
3 |
-
import
|
|
|
4 |
from io import BytesIO
|
|
|
|
|
|
|
5 |
import torch
|
6 |
from torchvision import transforms
|
7 |
from transformers import AutoModelForImageSegmentation
|
@@ -70,14 +74,15 @@ usage_to_weights_file = {
|
|
70 |
'General-legacy': 'BiRefNet-legacy'
|
71 |
}
|
72 |
|
73 |
-
|
|
|
74 |
birefnet.to(device)
|
75 |
birefnet.eval()
|
76 |
|
77 |
# Set resolution
|
78 |
-
if
|
79 |
resolution = (2560, 1440)
|
80 |
-
elif
|
81 |
resolution = (512, 512)
|
82 |
else:
|
83 |
resolution = (1024, 1024)
|
|
|
1 |
# These HF deployment codes refer to https://huggingface.co/not-lain/BiRefNet/raw/main/handler.py.
|
2 |
from typing import Dict, List, Any, Tuple
|
3 |
+
import os
|
4 |
+
import requests
|
5 |
from io import BytesIO
|
6 |
+
import cv2
|
7 |
+
import numpy as np
|
8 |
+
from PIL import Image
|
9 |
import torch
|
10 |
from torchvision import transforms
|
11 |
from transformers import AutoModelForImageSegmentation
|
|
|
74 |
'General-legacy': 'BiRefNet-legacy'
|
75 |
}
|
76 |
|
77 |
+
usage = 'General'
|
78 |
+
birefnet = AutoModelForImageSegmentation.from_pretrained('/'.join(('zhengpeng7', usage_to_weights_file[usage])), trust_remote_code=True)
|
79 |
birefnet.to(device)
|
80 |
birefnet.eval()
|
81 |
|
82 |
# Set resolution
|
83 |
+
if usage in ['General-Lite-2K']:
|
84 |
resolution = (2560, 1440)
|
85 |
+
elif usage in ['General-reso_512']:
|
86 |
resolution = (512, 512)
|
87 |
else:
|
88 |
resolution = (1024, 1024)
|