File size: 10,147 Bytes
8e0bf0b
 
 
 
9c130b3
8e0bf0b
 
 
 
 
 
 
 
4ea140d
 
2ce2894
2186adb
2f12544
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c32d0ce
fe9aabc
 
 
 
5610b71
fe9aabc
5610b71
fe9aabc
 
 
 
 
5610b71
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8e0bf0b
 
 
 
4e4a450
 
aa7a82a
 
904d41c
 
 
 
 
cfc0ae6
9f7b69e
8e0bf0b
279855a
3324530
4e4a450
8e0bf0b
 
 
 
 
 
 
78a1cfe
564a287
8e6daa5
 
3be73e9
fe9aabc
 
 
74f851f
fe9aabc
 
8e0bf0b
 
 
6ce55a1
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
326
import functools
import os
import shutil
import sys
import git

import gradio as gr
import numpy as np
import torch as torch
from PIL import Image

from gradio_imageslider import ImageSlider

import spaces

@spaces.GPU
def depth_normal(img):

    pipe_out = pipe(
        input_image,
        denoising_steps=10,
        ensemble_size=1,
        processing_res=768,
        batch_size=0,
        guidance_scale=3,
        domain="indoor",
        show_progress_bar=True,
    )

    depth_colored = pipe_out.depth_colored
    normal_colored = pipe_out.normal_colored
    
    return depth_colored, normal_colored

# @spaces.GPU
# def run_demo_server(pipe):
#     title = "Geowizard"
#     description = "Gradio demo for Geowizard."

#     examples = ["files/bee.jpg"]
    
#     # gr.Interface(
#     #     depth_normal, 
#     #     inputs=[gr.Image(type='pil', label="Original Image")], 
#     #     outputs=[gr.Image(type="pil",label="Output Depth"), gr.Image(type="pil",label="Output Normal")], 
#     #     title=title, description=description, article='1', examples=examples, analytics_enabled=False).launch()

    
# def process(
#     pipe,
#     path_input,
#     ensemble_size,
#     denoise_steps,
#     processing_res,
#     path_out_16bit=None,
#     path_out_fp32=None,
#     path_out_vis=None,
# ):

#     if path_out_vis is not None:
#         return (
#             [path_out_16bit, path_out_vis],
#             [path_out_16bit, path_out_fp32, path_out_vis],
#         )

#     input_image = Image.open(path_input)

#     pipe_out = pipe(
#         input_image,
#         denoising_steps=denoise_steps,
#         ensemble_size=ensemble_size,
#         processing_res=processing_res,
#         batch_size=1 if processing_res == 0 else 0,
#         guidance_scale=3,
#         domain="indoor",
#         show_progress_bar=True,
#     )

#     depth_pred = pipe_out.depth_np
#     depth_colored = pipe_out.depth_colored
#     depth_16bit = (depth_pred * 65535.0).astype(np.uint16)

#     path_output_dir = os.path.splitext(path_input)[0] + "_output"
#     os.makedirs(path_output_dir, exist_ok=True)

#     name_base = os.path.splitext(os.path.basename(path_input))[0]
#     path_out_fp32 = os.path.join(path_output_dir, f"{name_base}_depth_fp32.npy")
#     path_out_16bit = os.path.join(path_output_dir, f"{name_base}_depth_16bit.png")
#     path_out_vis = os.path.join(path_output_dir, f"{name_base}_depth_colored.png")

#     np.save(path_out_fp32, depth_pred)
#     Image.fromarray(depth_16bit).save(path_out_16bit, mode="I;16")
#     depth_colored.save(path_out_vis)

#     return (
#         [path_out_16bit, path_out_vis],
#         [path_out_16bit, path_out_fp32, path_out_vis],
#     )


# @spaces.GPU
# def run_demo_server(pipe):
#     process_pipe = functools.partial(process, pipe)
#     os.environ["GRADIO_ALLOW_FLAGGING"] = "never"

#     with gr.Blocks(
#         analytics_enabled=False,
#         title="GeoWizard Depth and Normal Estimation",
#         css="""
#             #download {
#                 height: 118px;
#             }
#             .slider .inner {
#                 width: 5px;
#                 background: #FFF;
#             }
#             .viewport {
#                 aspect-ratio: 4/3;
#             }
#         """,
#     ) as demo:
#         gr.Markdown(
#         """
#             <h1 align="center">Geowizard Depth & Normal Estimation</h1>
#         """
#         )

#         with gr.Row():
#             with gr.Column():
#                 input_image = gr.Image(
#                     label="Input Image",
#                     type="filepath",
#                 )
#                 with gr.Accordion("Advanced options", open=False):
#                     domain = gr.Radio(
#                         [
#                             ("Outdoor", "outdoor"),
#                             ("Indoor", "indoor"),
#                             ("Object", "object"),
#                         ],
#                         label="Data Domain",
#                         value="indoor",
#                     )
#                     cfg_scale = gr.Slider(
#                         label="Classifier Free Guidance Scale",
#                         minimum=1,
#                         maximum=5,
#                         step=1,
#                         value=3,
#                     )
#                     denoise_steps = gr.Slider(
#                         label="Number of denoising steps",
#                         minimum=1,
#                         maximum=20,
#                         step=1,
#                         value=2,
#                     )
#                     ensemble_size = gr.Slider(
#                         label="Ensemble size",
#                         minimum=1,
#                         maximum=15,
#                         step=1,
#                         value=1,
#                     )
#                     processing_res = gr.Radio(
#                         [
#                             ("Native", 0),
#                             ("Recommended", 768),
#                         ],
#                         label="Processing resolution",
#                         value=768,
#                     )
#                 input_output_16bit = gr.File(
#                     label="Predicted depth (16-bit)",
#                     visible=False,
#                 )
#                 input_output_fp32 = gr.File(
#                     label="Predicted depth (32-bit)",
#                     visible=False,
#                 )
#                 input_output_vis = gr.File(
#                     label="Predicted depth (red-near, blue-far)",
#                     visible=False,
#                 )
#                 with gr.Row():
#                     submit_btn = gr.Button(value="Compute", variant="primary")
#                     clear_btn = gr.Button(value="Clear")
#             with gr.Column():
#                 output_slider = ImageSlider(
#                     label="Predicted depth (red-near, blue-far)",
#                     type="filepath",
#                     show_download_button=True,
#                     show_share_button=True,
#                     interactive=False,
#                     elem_classes="slider",
#                     position=0.25,
#                 )
#                 files = gr.Files(
#                     label="Depth outputs",
#                     elem_id="download",
#                     interactive=False,
#                 )

#         blocks_settings_depth = [ensemble_size, denoise_steps, processing_res]
#         blocks_settings = blocks_settings_depth
#         map_id_to_default = {b._id: b.value for b in blocks_settings}

#         inputs = [
#             input_image,
#             ensemble_size,
#             denoise_steps,
#             processing_res,
#             input_output_16bit,
#             input_output_fp32,
#             input_output_vis,
#         ]
#         outputs = [
#             submit_btn,
#             input_image,
#             output_slider,
#             files,
#         ]

#         def submit_depth_fn(*args):
#             out = list(process_pipe(*args))
#             out = [gr.Button(interactive=False), gr.Image(interactive=False)] + out
#             return out

#         submit_btn.click(
#             fn=submit_depth_fn,
#             inputs=inputs,
#             outputs=outputs,
#             concurrency_limit=1,
#         )

#         gr.Examples(
#             fn=submit_depth_fn,
#             examples=[
#                 [
#                     "files/bee.jpg",
#                     10,  # ensemble_size
#                     10,  # denoise_steps
#                     768,  # processing_res
#                     "files/bee_depth_16bit.png",
#                     "files/bee_depth_fp32.npy",
#                     "files/bee_depth_colored.png",
#                 ],
#             ],
#             inputs=inputs,
#             outputs=outputs,
#             cache_examples=True,
#         )

#         def clear_fn():
#             out = []
#             for b in blocks_settings:
#                 out.append(map_id_to_default[b._id])
#             out += [
#                 gr.Button(interactive=True),
#                 gr.Image(value=None, interactive=True),
#                 None, None, None, None, None, None, None,
#             ]
#             return out

#         clear_btn.click(
#             fn=clear_fn,
#             inputs=[],
#             outputs=blocks_settings + [
#                 submit_btn,
#                 input_image,
#                 input_output_16bit,
#                 input_output_fp32,
#                 input_output_vis,
#                 output_slider,
#                 files,
#             ],
#         )

#         demo.queue(
#             api_open=False,
#         ).launch(
#             server_name="0.0.0.0",
#             server_port=7860,
#         )


def main():

    REPO_URL = "https://github.com/lemonaddie/geowizard.git"
    CHECKPOINT = "lemonaddie/Geowizard"
    REPO_DIR = "geowizard"
    
    if os.path.isdir(REPO_DIR):
        shutil.rmtree(REPO_DIR)
    
    repo = git.Repo.clone_from(REPO_URL, REPO_DIR)
    sys.path.append(os.path.join(os.getcwd(), REPO_DIR))

    from pipeline.depth_normal_pipeline_clip_cfg import DepthNormalEstimationPipeline

    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")  
    pipe = DepthNormalEstimationPipeline.from_pretrained(CHECKPOINT)
    
    try:
        import xformers
        pipe.enable_xformers_memory_efficient_attention()
    except:
        pass  # run without xformers

    pipe = pipe.to(device)
    #run_demo_server(pipe)

    title = "Geowizard"
    description = "Gradio demo for Geowizard."
    examples = ["files/bee.jpg"]

    gr.Interface(
        depth_normal, 
        inputs=[gr.Image(type='pil', label="Original Image")], 
        outputs=[gr.Image(type="pil",label="Output Depth"), gr.Image(type="pil",label="Output Normal")], 
        title=title, description=description, article='1', examples=examples, analytics_enabled=False).launch()


if __name__ == "__main__":
    main()