File size: 5,688 Bytes
ec0fdfd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import argparse


def get_arguments() -> argparse.Namespace:
    parser = argparse.ArgumentParser()

    # Inputs
    parser.add_argument(
        "-p", "--prompt", type=str, help="The prompt for the desired editing", required=True
    )
    parser.add_argument(
        "-i", "--init_image", type=str, help="The path to the source image input", required=True
    )
    parser.add_argument(
        "-i2", "--init_image_2", type=str, help="The path to the source image input", required=True
    )

    parser.add_argument("--mask", type=str, help="The path to the mask to edit with", default=None)

    # Diffusion
    parser.add_argument(
        "--skip_timesteps",
        type=int,
        help="How many steps to skip during the diffusion.",
        default=25,
    )
    parser.add_argument(
        "--local_clip_guided_diffusion",
        help="Indicator for using local CLIP guided diffusion (for baseline comparison)",
        action="store_true",
        dest="local_clip_guided_diffusion",
    )
    parser.add_argument(
        "--ddim",
        help="Indicator for using DDIM instead of DDPM",
        action="store_true",
    )

    # For more details read guided-diffusion/guided_diffusion/respace.py
    parser.add_argument(
        "--timestep_respacing",
        type=str,
        help="How to respace the intervals of the diffusion process (number between 1 and 1000).",
        default="100",
    )
    parser.add_argument(
        "--model_output_size",
        type=int,
        help="The resolution of the outputs of the diffusion model",
        default=256,
        choices=[256, 512],
    )

    # Augmentations
    parser.add_argument("--aug_num", type=int, help="The number of augmentation", default=8)

    # Loss
    parser.add_argument(
        "--clip_guidance_lambda",
        type=float,
        help="Controls how much the image should look like the prompt",
        default=1000,
    )
    parser.add_argument(
        "--range_lambda",
        type=float,
        help="Controls how far out of range RGB values are allowed to be",
        default=50,
    )
    parser.add_argument(
        "--lpips_sim_lambda",
        type=float,
        help="The LPIPS similarity to the input image",
        default=1000,
    )
    parser.add_argument(
        "--l2_sim_lambda", type=float, help="The L2 similarity to the input image", default=10000,
    )
    parser.add_argument(
        "--background_preservation_loss",
        help="Indicator for using the background preservation loss",
        action="store_true",
    )

    # Mask
    parser.add_argument(
        "--invert_mask",
        help="Indicator for mask inversion",
        action="store_true",
        dest="invert_mask",
    )
    parser.add_argument(
        "--no_enforce_background",
        help="Indicator disabling the last background enforcement",
        action="store_false",
        dest="enforce_background",
    )

    # Misc
    parser.add_argument("--seed", type=int, help="The random seed", default=404)
    parser.add_argument("--gpu_id", type=int, help="The GPU ID", default=0)
    parser.add_argument("--output_path", type=str, default="output")
    parser.add_argument(
        "-o",
        "--output_file",
        type=str,
        help="The filename to save, must be png",
        default="output.png",
    )
    parser.add_argument("--iterations_num", type=int, help="The number of iterations", default=8)
    parser.add_argument(
        "--batch_size",
        type=int,
        help="The number number if images to sample each diffusion process",
        default=4,
    )
    parser.add_argument(
        "--vid",
        help="Indicator for saving the video of the diffusion process",
        action="store_true",
        dest="save_video",
    )
    parser.add_argument(
        "--export_assets",
        help="Indicator for saving raw assets of the prediction",
        action="store_true",
        dest="export_assets",
    )
    parser.add_argument(
        "--image_guide",
        help="Indicator image or text",
        action="store_true",
        dest="image_guide",
    )
    parser.add_argument(
        "--coarse_to_fine",
        help="Indicator mask from big to small",
        action="store_true",
        dest="coarse_to_fine",
    )
    parser.add_argument(
        "--classifier_scale",
        type=float,
        help="Classifer scale for class guided",
        default=10.,
    )
    parser.add_argument(
        "--y",
        type=int,
        help="Target class for classifier guidence",
        default=0,
    )
    parser.add_argument(
        "--class_cond",
        help="classifer conditioned for diffusion model or not",
        action="store_true",
        dest="class_cond",
    )
    parser.add_argument(
        "--background_complex",
        type=float,
        help="BG complex guidance scale",
        default=0.,
    )
    parser.add_argument(
        "--final_save_root",
        type=str,
        help="Final save root",
        default="validation-generated/generated-with-25-steps-bg/final/",
    )
    parser.add_argument(
        "--hard",
        help="hard or smooth",
        action="store_true",
        dest="hard",
    )
    parser.add_argument(
        "--random_position",
        help="apply random position",
        action="store_true",
        dest="random_position",
    )
    parser.add_argument(
        "--rotate_obj",
        help="apply random rotate to objects",
        action="store_true",
        dest="rotate_obj",
    )
    parser.add_argument(
        "--angle",
        type=int,
        help="angle",
        default=0,
    )
    args = parser.parse_args()
    print(args)
    return args