File size: 5,950 Bytes
035e10c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import torch
import torch.nn as nn
from torch.nn import init
from torch.optim import lr_scheduler


def get_scheduler(optimizer, opt):
    if opt.lr_policy == 'linear':
        def lambda_rule(epoch):
            # lr_l = 1.0 - max(0, epoch + opt.epoch_count - opt.n_epochs) / float(opt.n_epochs_decay + 1)
            lr_l = 0.3 ** max(0, (epoch + opt.epoch_count - opt.n_epochs) // 5)
            return lr_l

        scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lambda_rule)
    elif opt.lr_policy == 'step':
        scheduler = lr_scheduler.StepLR(optimizer, step_size=opt.lr_decay_iters, gamma=0.1)
    elif opt.lr_policy == 'plateau':
        scheduler = lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', factor=0.2, threshold=0.01, patience=5)
    elif opt.lr_policy == 'cosine':
        scheduler = lr_scheduler.CosineAnnealingLR(optimizer, T_max=opt.n_epochs, eta_min=0)
    else:
        return NotImplementedError('learning rate policy [%s] is not implemented', opt.lr_policy)
    return scheduler


def init_weights(net, init_type='normal', init_gain=0.02):
    def init_func(m):
        classname = m.__class__.__name__
        if hasattr(m, 'weight') and (classname.find('Conv') != -1 or classname.find('Linear') != -1):
            if init_type == 'normal':
                init.normal_(m.weight.data, 0.0, init_gain)
            elif init_type == 'xavier':
                init.xavier_normal_(m.weight.data, gain=init_gain)
            elif init_type == 'kaiming':
                init.kaiming_normal_(m.weight.data, a=0, mode='fan_in')
            elif init_type == 'orthogonal':
                init.orthogonal_(m.weight.data, gain=init_gain)
            else:
                raise NotImplementedError('initialization method [%s] is not implemented' % init_type)
            if hasattr(m, 'bias') and m.bias is not None:
                init.constant_(m.bias.data, 0.0)
        elif classname.find('BatchNorm2d') != -1:
            init.normal_(m.weight.data, 1.0, init_gain)
            init.constant_(m.bias.data, 0.0)

    print('initialize network with %s' % init_type)
    net.apply(init_func)


def init_net(net, init_type='normal', init_gain=0.02, gpu_ids=()):
    if len(gpu_ids) > 0:
        assert (torch.cuda.is_available())
        net.to(gpu_ids[0])
        net = torch.nn.DataParallel(net, gpu_ids)  # multi-GPUs
    init_weights(net, init_type, init_gain=init_gain)
    return net


class SignWithSigmoidGrad(torch.autograd.Function):

    @staticmethod
    def forward(ctx, x):
        result = (x > 0).float()
        sigmoid_result = torch.sigmoid(x)
        ctx.save_for_backward(sigmoid_result)
        return result

    @staticmethod
    def backward(ctx, grad_result):
        (sigmoid_result,) = ctx.saved_tensors
        if ctx.needs_input_grad[0]:
            grad_input = grad_result * sigmoid_result * (1 - sigmoid_result)
        else:
            grad_input = None
        return grad_input


class Painter(nn.Module):

    def __init__(self, param_per_stroke, total_strokes, hidden_dim, n_heads=8, n_enc_layers=3, n_dec_layers=3):
        super().__init__()
        self.enc_img = nn.Sequential(
            nn.ReflectionPad2d(1),
            nn.Conv2d(3, 32, 3, 1),
            nn.BatchNorm2d(32),
            nn.ReLU(True),
            nn.ReflectionPad2d(1),
            nn.Conv2d(32, 64, 3, 2),
            nn.BatchNorm2d(64),
            nn.ReLU(True),
            nn.ReflectionPad2d(1),
            nn.Conv2d(64, 128, 3, 2),
            nn.BatchNorm2d(128),
            nn.ReLU(True))
        self.enc_canvas = nn.Sequential(
            nn.ReflectionPad2d(1),
            nn.Conv2d(3, 32, 3, 1),
            nn.BatchNorm2d(32),
            nn.ReLU(True),
            nn.ReflectionPad2d(1),
            nn.Conv2d(32, 64, 3, 2),
            nn.BatchNorm2d(64),
            nn.ReLU(True),
            nn.ReflectionPad2d(1),
            nn.Conv2d(64, 128, 3, 2),
            nn.BatchNorm2d(128),
            nn.ReLU(True))
        self.conv = nn.Conv2d(128 * 2, hidden_dim, 1)
        self.transformer = nn.Transformer(hidden_dim, n_heads, n_enc_layers, n_dec_layers)
        self.linear_param = nn.Sequential(
            nn.Linear(hidden_dim, hidden_dim),
            nn.ReLU(True),
            nn.Linear(hidden_dim, hidden_dim),
            nn.ReLU(True),
            nn.Linear(hidden_dim, param_per_stroke))
        self.linear_decider = nn.Linear(hidden_dim, 1)
        self.query_pos = nn.Parameter(torch.rand(total_strokes, hidden_dim))
        self.row_embed = nn.Parameter(torch.rand(8, hidden_dim // 2))
        self.col_embed = nn.Parameter(torch.rand(8, hidden_dim // 2))

    def forward(self, img, canvas):
        b, _, H, W = img.shape
        img_feat = self.enc_img(img)
        canvas_feat = self.enc_canvas(canvas)
        h, w = img_feat.shape[-2:]
        feat = torch.cat([img_feat, canvas_feat], dim=1)
        feat_conv = self.conv(feat)

        pos_embed = torch.cat([
            self.col_embed[:w].unsqueeze(0).contiguous().repeat(h, 1, 1),
            self.row_embed[:h].unsqueeze(1).contiguous().repeat(1, w, 1),
        ], dim=-1).flatten(0, 1).unsqueeze(1)
        hidden_state = self.transformer(pos_embed + feat_conv.flatten(2).permute(2, 0, 1).contiguous(),
                                        self.query_pos.unsqueeze(1).contiguous().repeat(1, b, 1))
        hidden_state = hidden_state.permute(1, 0, 2).contiguous()
        param = self.linear_param(hidden_state)
        s = hidden_state.shape[1]
        grid = param[:, :, :2].view(b * s, 1, 1, 2).contiguous()
        img_temp = img.unsqueeze(1).contiguous().repeat(1, s, 1, 1, 1).view(b * s, 3, H, W).contiguous()
        color = nn.functional.grid_sample(img_temp, 2 * grid - 1, align_corners=False).view(b, s, 3).contiguous()
        decision = self.linear_decider(hidden_state)
        return torch.cat([param, color, color, torch.rand(b, s, 1, device=img.device)], dim=-1), decision