File size: 5,791 Bytes
29d49a9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
# This is an improved version and model of HED edge detection with Apache License, Version 2.0.
# Please use this implementation in your products
# This implementation may produce slightly different results from Saining Xie's official implementations,
# but it generates smoother edges and is more suitable for ControlNet as well as other image-to-image translations.
# Different from official models and other implementations, this is an RGB-input model (rather than BGR)
# and in this way it works better for gradio's RGB protocol

import os
import warnings

import cv2
import numpy as np
import torch
from einops import rearrange
from huggingface_hub import hf_hub_download
from PIL import Image

from ..util import HWC3, nms, resize_image, safe_step


class DoubleConvBlock(torch.nn.Module):
    def __init__(self, input_channel, output_channel, layer_number):
        super().__init__()
        self.convs = torch.nn.Sequential()
        self.convs.append(torch.nn.Conv2d(in_channels=input_channel, out_channels=output_channel, kernel_size=(3, 3), stride=(1, 1), padding=1))
        for i in range(1, layer_number):
            self.convs.append(torch.nn.Conv2d(in_channels=output_channel, out_channels=output_channel, kernel_size=(3, 3), stride=(1, 1), padding=1))
        self.projection = torch.nn.Conv2d(in_channels=output_channel, out_channels=1, kernel_size=(1, 1), stride=(1, 1), padding=0)

    def __call__(self, x, down_sampling=False):
        h = x
        if down_sampling:
            h = torch.nn.functional.max_pool2d(h, kernel_size=(2, 2), stride=(2, 2))
        for conv in self.convs:
            h = conv(h)
            h = torch.nn.functional.relu(h)
        return h, self.projection(h)


class ControlNetHED_Apache2(torch.nn.Module):
    def __init__(self):
        super().__init__()
        self.norm = torch.nn.Parameter(torch.zeros(size=(1, 3, 1, 1)))
        self.block1 = DoubleConvBlock(input_channel=3, output_channel=64, layer_number=2)
        self.block2 = DoubleConvBlock(input_channel=64, output_channel=128, layer_number=2)
        self.block3 = DoubleConvBlock(input_channel=128, output_channel=256, layer_number=3)
        self.block4 = DoubleConvBlock(input_channel=256, output_channel=512, layer_number=3)
        self.block5 = DoubleConvBlock(input_channel=512, output_channel=512, layer_number=3)

    def __call__(self, x):
        h = x - self.norm
        h, projection1 = self.block1(h)
        h, projection2 = self.block2(h, down_sampling=True)
        h, projection3 = self.block3(h, down_sampling=True)
        h, projection4 = self.block4(h, down_sampling=True)
        h, projection5 = self.block5(h, down_sampling=True)
        return projection1, projection2, projection3, projection4, projection5

class HEDdetector:
    def __init__(self, netNetwork):
        self.netNetwork = netNetwork

    @classmethod
    def from_pretrained(cls, pretrained_model_or_path, filename=None, cache_dir=None, local_files_only=False):
        filename = filename or "ControlNetHED.pth"

        if os.path.isdir(pretrained_model_or_path):
            model_path = os.path.join(pretrained_model_or_path, filename)
        else:
            model_path = hf_hub_download(pretrained_model_or_path, filename, cache_dir=cache_dir, local_files_only=local_files_only)

        netNetwork = ControlNetHED_Apache2()
        netNetwork.load_state_dict(torch.load(model_path, map_location='cpu'))
        netNetwork.float().eval()

        return cls(netNetwork)
    
    def to(self, device):
        self.netNetwork.to(device)
        return self
    
    def __call__(self, input_image, detect_resolution=512, image_resolution=512, safe=False, output_type="pil", scribble=False, **kwargs):
        if "return_pil" in kwargs:
            warnings.warn("return_pil is deprecated. Use output_type instead.", DeprecationWarning)
            output_type = "pil" if kwargs["return_pil"] else "np"
        if type(output_type) is bool:
            warnings.warn("Passing `True` or `False` to `output_type` is deprecated and will raise an error in future versions")
            if output_type:
                output_type = "pil"

        device = next(iter(self.netNetwork.parameters())).device
        if not isinstance(input_image, np.ndarray):
            input_image = np.array(input_image, dtype=np.uint8)

        input_image = HWC3(input_image)
        input_image = resize_image(input_image, detect_resolution)

        assert input_image.ndim == 3
        H, W, C = input_image.shape
        with torch.no_grad():
            image_hed = torch.from_numpy(input_image.copy()).float().to(device)
            image_hed = rearrange(image_hed, 'h w c -> 1 c h w')
            edges = self.netNetwork(image_hed)
            edges = [e.detach().cpu().numpy().astype(np.float32)[0, 0] for e in edges]
            edges = [cv2.resize(e, (W, H), interpolation=cv2.INTER_LINEAR) for e in edges]
            edges = np.stack(edges, axis=2)
            edge = 1 / (1 + np.exp(-np.mean(edges, axis=2).astype(np.float64)))
            if safe:
                edge = safe_step(edge)
            edge = (edge * 255.0).clip(0, 255).astype(np.uint8)

        detected_map = edge
        detected_map = HWC3(detected_map)

        img = resize_image(input_image, image_resolution)
        H, W, C = img.shape

        detected_map = cv2.resize(detected_map, (W, H), interpolation=cv2.INTER_LINEAR)
        
        if scribble:
            detected_map = nms(detected_map, 127, 3.0)
            detected_map = cv2.GaussianBlur(detected_map, (0, 0), 3.0)
            detected_map[detected_map > 4] = 255
            detected_map[detected_map < 255] = 0

        if output_type == "pil":
            detected_map = Image.fromarray(detected_map)

        return detected_map