Datasets:
slz1
/

ArXiv:
File size: 11,952 Bytes
159572f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
import torch
import torch.nn as nn
import numpy as np
import torchvision.transforms.functional as F

from PIL import Image
from typing import Optional
from functools import partial
from torch import Tensor
from torchvision import transforms

# from canny_tools import Canny  # canny edge detection
# from mmengine.hub import get_model  # segmentation
from transformers import DPTForDepthEstimation  # depth estimation

# from mmseg.models.losses.silog_loss import silog_loss
from torchvision.transforms import RandomCrop


def get_reward_model(task='segmentation', model_path='mmseg::upernet/upernet_r50_4xb4-160k_ade20k-512x512.py'):
    """Return reward model for different tasks.

    Args:
        task (str, optional): Task name. Defaults to 'segmentation'.
        model_path (str, optional): Model name or pre-trained path.

    """
    if task == 'segmentation':
        return get_model(model_path, pretrained=True)
    elif task == 'canny':
        return Canny()
    elif task == 'depth':
        return DPTForDepthEstimation.from_pretrained(model_path)
    elif task == 'lineart':
        model = LineDrawingModel()
        model.load_state_dict(torch.hub.load_state_dict_from_url(model_path, map_location=torch.device('cpu')))
        return model
    elif task == 'hed':
        return HEDdetector(model_path)
    else:
        raise not NotImplementedError("Only support segmentation, canny and depth for now.")


def get_reward_loss(predictions, labels, task='segmentation', **args):
    """Return reward loss for different tasks.

    Args:
        task (str, optional): Task name.

    Returns:
        torch.nn.Module: Loss class.
    """
    if task == 'segmentation':
        return nn.functional.cross_entropy(predictions, labels, ignore_index=255, **args)
    elif task == 'canny':
        loss = nn.functional.mse_loss(predictions, labels, **args).mean(2)
        return loss.mean((-1,-2))
    elif task in ['depth', 'lineart', 'hed']:
        loss = nn.functional.mse_loss(predictions, labels, **args)
        return loss
    else:
        raise not NotImplementedError("Only support segmentation, canny and depth for now.")


def image_grid(imgs, rows, cols):
    """Image grid for visualization."""
    assert len(imgs) == rows * cols

    w, h = imgs[0].size
    grid = Image.new("RGB", size=(cols * w, rows * h))

    for i, img in enumerate(imgs):
        grid.paste(img, box=(i % cols * w, i // cols * h))
    return grid


def map_color_to_index(image, dataset='limingcv/Captioned_ADE20K'):
    """Map colored segmentation image (RGB) into original label format (L).

    Args:
        image (torch.tensor): image tensor with shape (N, 3, H, W).
        dataset (str, optional): Dataset name. Defaults to 'ADE20K'.

    Returns:
        torch.tensor: mask tensor with shape (N, H, W).
    """
    if dataset == 'limingcv/Captioned_ADE20K':
        palette = np.load('ade20k_palette.npy')
    elif dataset == 'limingcv/Captioned_COCOStuff':
        palette = np.load('coco_stuff_palette.npy')
    else:
        raise NotImplementedError("Only support ADE20K and COCO-Stuff dataset for now.")

    image = image * 255
    palette_tensor = torch.tensor(palette, dtype=image.dtype, device=image.device)
    reshaped_image = image.permute(0, 2, 3, 1).reshape(-1, 3)

    # Calculate the difference of colors and find the index of the minimum distance
    indices = torch.argmin(torch.norm(reshaped_image[:, None, :] - palette_tensor, dim=-1), dim=-1)

    # Transform indices back to original shape
    return indices.view(image.shape[0], image.shape[2], image.shape[3])


def seg_label_transform(
        labels,
        dataset_name='limingcv/Captioned_ADE20K',
        output_size=(64, 64),
        interpolation=transforms.InterpolationMode.NEAREST,
        max_size=None,
        antialias=True):
    """Adapt RGB seg_map into loss computation. \
    (1) Map the RGB seg_map into the original label format (Single Channel). \
    (2) Resize the seg_map into the same size as the output feature map. \
    (3) Remove background class if needed (usually for ADE20K).

    Args:
        labels (torch.tensor): Segmentation map. (N, 3, H, W) for ADE20K and (N, H, W) for COCO-Stuff.
        dataset_name (string): Dataset name. Default to 'ADE20K'.
        output_size (tuple): Resized image size, should be aligned with the output of segmentation models.
        interpolation (optional): _description_. Defaults to transforms.InterpolationMode.NEAREST.
        max_size (optional): Defaults to None.
        antialias (optional): Defaults to True.

    Returns:
        torch.tensor: formatted labels for loss computation.
    """

    if dataset_name == 'limingcv/Captioned_ADE20K':
        labels = map_color_to_index(labels, dataset_name)
        labels = F.resize(labels, output_size, interpolation, max_size, antialias)

        # 0 means the background class in ADE20K
        # In a unified format, we use 255 to represent the background class for both ADE20K and COCO-Stuff
        labels = labels - 1
        labels[labels == -1] = 255
    elif dataset_name == 'limingcv/Captioned_COCOStuff':
        labels = F.resize(labels, output_size, interpolation, max_size, antialias)

    return labels.long()

def depth_label_transform(
        labels,
        dataset_name,
        output_size=None,
        interpolation=transforms.InterpolationMode.BILINEAR,
        max_size=None,
        antialias=True
    ):

    if output_size is not None:
        labels = F.resize(labels, output_size, interpolation, max_size, antialias)
    return labels


def edge_label_transform(labels, dataset_name):
    return labels


def label_transform(labels, task, dataset_name, **args):
    if task == 'segmentation':
        return seg_label_transform(labels, dataset_name, **args)
    elif task == 'depth':
        return depth_label_transform(labels, dataset_name, **args)
    elif task in ['canny', 'lineart', 'hed']:
        return edge_label_transform(labels, dataset_name, **args)
    else:
        raise NotImplementedError("Only support segmentation and edge detection for now.")


def group_random_crop(images, resolution):
    """
    Args:
        images (list of PIL Image or Tensor): List of images to be cropped.

    Returns:
        List of PIL Image or Tensor: List of cropped image.
    """

    if isinstance(resolution, int):
        resolution = (resolution, resolution)

    for idx, image in enumerate(images):
        i, j, h, w = RandomCrop.get_params(image, output_size=resolution)
        images[idx] = F.crop(image, i, j, h, w)

    return images


norm_layer = nn.InstanceNorm2d
class ResidualBlock(nn.Module):
    def __init__(self, in_features):
        super(ResidualBlock, self).__init__()

        conv_block = [  nn.ReflectionPad2d(1),
                        nn.Conv2d(in_features, in_features, 3),
                        norm_layer(in_features),
                        nn.ReLU(inplace=True),
                        nn.ReflectionPad2d(1),
                        nn.Conv2d(in_features, in_features, 3),
                        norm_layer(in_features)
                        ]

        self.conv_block = nn.Sequential(*conv_block)

    def forward(self, x):
        return x + self.conv_block(x)


class LineDrawingModel(nn.Module):
    def __init__(self, input_nc=3, output_nc=1, n_residual_blocks=3, sigmoid=True):
        super(LineDrawingModel, self).__init__()

        # Initial convolution block
        model0 = [   nn.ReflectionPad2d(3),
                    nn.Conv2d(input_nc, 64, 7),
                    norm_layer(64),
                    nn.ReLU(inplace=True) ]
        self.model0 = nn.Sequential(*model0)

        # Downsampling
        model1 = []
        in_features = 64
        out_features = in_features*2
        for _ in range(2):
            model1 += [  nn.Conv2d(in_features, out_features, 3, stride=2, padding=1),
                        norm_layer(out_features),
                        nn.ReLU(inplace=True) ]
            in_features = out_features
            out_features = in_features*2
        self.model1 = nn.Sequential(*model1)

        model2 = []
        # Residual blocks
        for _ in range(n_residual_blocks):
            model2 += [ResidualBlock(in_features)]
        self.model2 = nn.Sequential(*model2)

        # Upsampling
        model3 = []
        out_features = in_features//2
        for _ in range(2):
            model3 += [  nn.ConvTranspose2d(in_features, out_features, 3, stride=2, padding=1, output_padding=1),
                        norm_layer(out_features),
                        nn.ReLU(inplace=True) ]
            in_features = out_features
            out_features = in_features//2
        self.model3 = nn.Sequential(*model3)

        # Output layer
        model4 = [  nn.ReflectionPad2d(3),
                        nn.Conv2d(64, output_nc, 7)]
        if sigmoid:
            model4 += [nn.Sigmoid()]

        self.model4 = nn.Sequential(*model4)

    def forward(self, x, cond=None):
        out = self.model0(x)
        out = self.model1(out)
        out = self.model2(out)
        out = self.model3(out)
        out = self.model4(out)

        return out



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(nn.Module):
    def __init__(self, model_path):
        super().__init__()
        state_dict = torch.hub.load_state_dict_from_url(model_path, map_location=torch.device('cpu'))

        self.netNetwork = ControlNetHED_Apache2()
        self.netNetwork.load_state_dict(state_dict)

    def __call__(self, input_image):
        H, W = input_image.shape[2], input_image.shape[3]

        edges = self.netNetwork((input_image * 255).clip(0, 255))
        edges = [torch.nn.functional.interpolate(edge, size=(H, W), mode='bilinear') for edge in edges]
        edges = torch.stack(edges, dim=1)
        edge = 1 / (1 + torch.exp(-torch.mean(edges, dim=1)))
        edge = (edge * 255.0).clip(0, 255).to(torch.uint8)

        return edge / 255.0