File size: 29,319 Bytes
399e1c1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
try:
    import spaces
except:
    pass

import os
import gradio as gr

import torch
from gradio_image_prompter import ImagePrompter
from sam2.sam2_image_predictor import SAM2ImagePredictor
from omegaconf import OmegaConf
from PIL import Image
import numpy as np
from copy import deepcopy
import cv2

import torch.nn.functional as F
import torchvision
from einops import rearrange
import tempfile

from objctrl_2_5d.utils.ui_utils import process_image, get_camera_pose, get_subject_points, get_points, undo_points, mask_image
from ZoeDepth.zoedepth.utils.misc import colorize

from cameractrl.inference import get_pipeline
from objctrl_2_5d.utils.examples import examples, sync_points

from objctrl_2_5d.utils.objmask_util import RT2Plucker, Unprojected, roll_with_ignore_multidim, dilate_mask_pytorch
from objctrl_2_5d.utils.filter_utils import get_freq_filter, freq_mix_3d


### Title and Description ###
#### Description ####
title = r"""<h1 align="center">ObjCtrl-2.5D: Training-free Object Control with Camera Poses</h1>"""
# subtitle = r"""<h2 align="center">Deployed on SVD Generation</h2>"""
important_link = r"""
<div align='center'>
 <a href='https://wzhouxiff.github.io/projects/MotionCtrl/assets/paper/MotionCtrl.pdf'>[Paper]</a>
&ensp; <a href='https://wzhouxiff.github.io/projects/MotionCtrl/'>[Project Page]</a>
&ensp; <a href='https://github.com/TencentARC/MotionCtrl'>[Code]</a>
</div>
"""

authors = r"""
<div align='center'>
 <a href='https://wzhouxiff.github.io/'>Zhouxia Wang</a>
&ensp; <a href='https://nirvanalan.github.io/'>Yushi Lan</a>
&ensp; <a href='https://shangchenzhou.com/'>Shanchen Zhou</a>
&ensp; <a href='https://www.mmlab-ntu.com/person/ccloy/index.html'>Chen Change Loy</a>
</div>
"""

affiliation = r"""
<div align='center'>
 <a href='https://www.mmlab-ntu.com/'>S-Lab, NTU Singapore</a>
</div>
"""

description = r"""
<b>Official Gradio demo</b> for <a href='https://github.com/TencentARC/MotionCtrl' target='_blank'><b>ObjCtrl-2.5D: Training-free Object Control with Camera Poses</b></a>.<br>
πŸ”₯ ObjCtrl2.5D enables object motion control in a I2V generated video via transforming 2D trajectories to 3D using depth, subsequently converting them into camera poses, 
thereby leveraging the exisitng camera motion control module for object motion control without requiring additional training.<br>
"""

article = r"""
If ObjCtrl2.5D is helpful, please help to ⭐ the <a href='https://github.com/TencentARC/MotionCtrl' target='_blank'>Github Repo</a>. Thanks! 
[![GitHub Stars](https://img.shields.io/github/stars/TencentARC%2FMotionCtrl
)](https://github.com/TencentARC/MotionCtrl)

---

πŸ“ **Citation**
<br>
If our work is useful for your research, please consider citing:
```bibtex
@inproceedings{wang2024motionctrl,
  title={Motionctrl: A unified and flexible motion controller for video generation},
  author={Wang, Zhouxia and Yuan, Ziyang and Wang, Xintao and Li, Yaowei and Chen, Tianshui and Xia, Menghan and Luo, Ping and Shan, Ying},
  booktitle={ACM SIGGRAPH 2024 Conference Papers},
  pages={1--11},
  year={2024}
}
```

πŸ“§ **Contact**
<br>
If you have any questions, please feel free to reach me out at <b>zhouzi1212@gmail.com</b>.

"""

# -------------- initialization --------------

CAMERA_MODE = ["Traj2Cam", "Rotate", "Clockwise", "Translate"]

# select the device for computation
if torch.cuda.is_available():
    device = torch.device("cuda")
elif torch.backends.mps.is_available():
    device = torch.device("mps")
else:
    device = torch.device("cpu")
print(f"using device: {device}")

# segmentation model
segmentor = SAM2ImagePredictor.from_pretrained("facebook/sam2-hiera-tiny", cache_dir="ckpt", device=device)

# depth model
d_model_NK = torch.hub.load('./ZoeDepth', 'ZoeD_NK', source='local', pretrained=True).to(device)

# cameractrl model
config = "configs/svd_320_576_cameractrl.yaml"
model_id = "stabilityai/stable-video-diffusion-img2vid"
ckpt = "checkpoints/CameraCtrl_svd.ckpt"
if not os.path.exists(ckpt):
    os.makedirs("checkpoints", exist_ok=True)
    os.system("wget -c https://huggingface.co/hehao13/CameraCtrl_SVD_ckpts/resolve/main/CameraCtrl_svd.ckpt?download=true")
    os.system("mv CameraCtrl_svd.ckpt?download=true checkpoints/CameraCtrl_svd.ckpt")
model_config = OmegaConf.load(config)


pipeline = get_pipeline(model_id, "unet", model_config['down_block_types'], model_config['up_block_types'],
                        model_config['pose_encoder_kwargs'], model_config['attention_processor_kwargs'],
                        ckpt, True, device)

# segmentor = None
# d_model_NK = None
# pipeline = None

### run the demo ##
# @spaces.GPU(duration=5)
def segment(canvas, image, logits):
    if logits is not None:
        logits *=  32.0
    _, points = get_subject_points(canvas)
    image = np.array(image)

    with torch.inference_mode(), torch.autocast("cuda", dtype=torch.bfloat16):
        segmentor.set_image(image)
        input_points = []
        input_boxes = []
        for p in points:
            [x1, y1, _, x2, y2, _] = p
            if x2==0 and y2==0:
                input_points.append([x1, y1])
            else:
                input_boxes.append([x1, y1, x2, y2])
        if len(input_points) == 0:
            input_points = None
            input_labels = None
        else:
            input_points = np.array(input_points)
            input_labels = np.ones(len(input_points))
        if len(input_boxes) == 0:
            input_boxes = None
        else:
            input_boxes = np.array(input_boxes)
        masks, _, logits = segmentor.predict(
            point_coords=input_points,
            point_labels=input_labels,
            box=input_boxes,
            multimask_output=False,
            return_logits=True,
            mask_input=logits,
        )
        mask = masks > 0
        masked_img = mask_image(image, mask[0], color=[252, 140, 90], alpha=0.9)
        masked_img = Image.fromarray(masked_img)
        
    return mask[0], masked_img, masked_img, logits / 32.0

# @spaces.GPU(duration=5)
def get_depth(image, points):
    
    depth = d_model_NK.infer_pil(image)    
    colored_depth = colorize(depth, cmap='gray_r') # [h, w, 4] 0-255
    
    depth_img = deepcopy(colored_depth[:, :, :3])
    if len(points) > 0:
        for idx, point in enumerate(points):
            if idx % 2 == 0:
                cv2.circle(depth_img, tuple(point), 10, (255, 0, 0), -1)
            else:
                cv2.circle(depth_img, tuple(point), 10, (0, 0, 255), -1)
            if idx > 0:
                cv2.arrowedLine(depth_img, points[idx-1], points[idx], (255, 255, 255), 4, tipLength=0.5)
    
    return depth, depth_img, colored_depth[:, :, :3]


# @spaces.GPU(duration=80)
def run_objctrl_2_5d(condition_image, 
                        mask, 
                        depth, 
                        RTs, 
                        bg_mode, 
                        shared_wapring_latents, 
                        scale_wise_masks, 
                        rescale, 
                        seed, 
                        ds, dt, 
                        num_inference_steps=25):
    
    DEBUG = False

    if DEBUG:
        cur_OUTPUT_PATH = 'outputs/tmp'
        os.makedirs(cur_OUTPUT_PATH, exist_ok=True)

    # num_inference_steps=25
    min_guidance_scale = 1.0
    max_guidance_scale = 3.0

    area_ratio = 0.3
    depth_scale_ = 5.2
    center_margin = 10

    height, width = 320, 576
    num_frames = 14

    intrinsics = np.array([[float(width), float(width), float(width) / 2, float(height) / 2]])
    intrinsics = np.repeat(intrinsics, num_frames, axis=0) # [n_frame, 4]
    fx = intrinsics[0, 0] / width
    fy = intrinsics[0, 1] / height
    cx = intrinsics[0, 2] / width
    cy = intrinsics[0, 3] / height

    down_scale = 8
    H, W = height // down_scale, width // down_scale
    K = np.array([[width / down_scale, 0, W / 2], [0, width / down_scale, H / 2], [0, 0, 1]])
    
    seed = int(seed)
            
    center_h_margin, center_w_margin = center_margin, center_margin
    depth_center = np.mean(depth[height//2-center_h_margin:height//2+center_h_margin, width//2-center_w_margin:width//2+center_w_margin])
    
    if rescale > 0:
        depth_rescale = round(depth_scale_ * rescale / depth_center, 2)
    else:
        depth_rescale = 1.0
        
    depth = depth * depth_rescale
    
    depth_down = F.interpolate(torch.tensor(depth).unsqueeze(0).unsqueeze(0), 
                                (H, W), mode='bilinear', align_corners=False).squeeze().numpy() # [H, W]
    
    ## latent
    generator = torch.Generator()
    generator.manual_seed(seed)
    
    latents_org = pipeline.prepare_latents(
            1,
            14,
            8,
            height,
            width,
            pipeline.dtype,
            device,
            generator,
            None,
        )
    latents_org = latents_org / pipeline.scheduler.init_noise_sigma
    
    cur_plucker_embedding, _, _ = RT2Plucker(RTs, RTs.shape[0], (height, width), fx, fy, cx, cy) # 6, V, H, W
    cur_plucker_embedding = cur_plucker_embedding.to(device)
    cur_plucker_embedding = cur_plucker_embedding[None, ...] # b 6 f h w
    cur_plucker_embedding = cur_plucker_embedding.permute(0, 2, 1, 3, 4) # b f 6 h w
    cur_plucker_embedding = cur_plucker_embedding[:, :num_frames, ...]
    cur_pose_features = pipeline.pose_encoder(cur_plucker_embedding)
    
    # bg_mode = ["Fixed", "Reverse", "Free"]
    if bg_mode == "Fixed":
        fix_RTs = np.repeat(RTs[0][None, ...], num_frames, axis=0) # [n_frame, 4, 3]
        fix_plucker_embedding, _, _ = RT2Plucker(fix_RTs, num_frames, (height, width), fx, fy, cx, cy) # 6, V, H, W
        fix_plucker_embedding = fix_plucker_embedding.to(device)
        fix_plucker_embedding = fix_plucker_embedding[None, ...] # b 6 f h w
        fix_plucker_embedding = fix_plucker_embedding.permute(0, 2, 1, 3, 4) # b f 6 h w
        fix_plucker_embedding = fix_plucker_embedding[:, :num_frames, ...]
        fix_pose_features = pipeline.pose_encoder(fix_plucker_embedding)
        
    elif bg_mode == "Reverse":
        bg_plucker_embedding, _, _ = RT2Plucker(RTs[::-1], RTs.shape[0], (height, width), fx, fy, cx, cy) # 6, V, H, W
        bg_plucker_embedding = bg_plucker_embedding.to(device)
        bg_plucker_embedding = bg_plucker_embedding[None, ...] # b 6 f h w
        bg_plucker_embedding = bg_plucker_embedding.permute(0, 2, 1, 3, 4) # b f 6 h w
        bg_plucker_embedding = bg_plucker_embedding[:, :num_frames, ...]
        fix_pose_features = pipeline.pose_encoder(bg_plucker_embedding)
        
    else:
        fix_pose_features = None
        
    #### preparing mask
    
    mask = Image.fromarray(mask)
    mask = mask.resize((W, H))
    mask = np.array(mask).astype(np.float32)
    mask = np.expand_dims(mask, axis=-1)
    
    # visulize mask
    if DEBUG:
        mask_sum_vis = mask[..., 0]
        mask_sum_vis = (mask_sum_vis * 255.0).astype(np.uint8)
        mask_sum_vis = Image.fromarray(mask_sum_vis)
        
        mask_sum_vis.save(f'{cur_OUTPUT_PATH}/org_mask.png')
    
    try:
        warped_masks = Unprojected(mask, depth_down, RTs, H=H, W=W, K=K)
    
        warped_masks.insert(0, mask)
            
    except:
        # mask to bbox
        print(f'!!! Mask is too small to warp; mask to bbox') 
        mask = mask[:, :, 0]
        coords = cv2.findNonZero(mask)
        x, y, w, h = cv2.boundingRect(coords)
        # mask[y:y+h, x:x+w] = 1.0
        
        center_x, center_y = x + w // 2, y + h // 2
        center_z = depth_down[center_y, center_x]
        
        # RTs [n_frame, 3, 4] to [n_frame, 4, 4] , add [0, 0, 0, 1]
        RTs = np.concatenate([RTs, np.array([[[0, 0, 0, 1]]] * num_frames)], axis=1)
        
        # RTs: world to camera
        P0 = np.array([center_x, center_y, 1])
        Pc0 = np.linalg.inv(K) @ P0 * center_z
        pw = np.linalg.inv(RTs[0]) @ np.array([Pc0[0], Pc0[1], center_z, 1]) # [4]
        
        P = [np.array([center_x, center_y])]
        for i in range(1, num_frames):
            Pci = RTs[i] @ pw
            Pi = K @ Pci[:3] / Pci[2]
            P.append(Pi[:2])
        
        warped_masks = [mask]
        for i in range(1, num_frames):
            shift_x = int(round(P[i][0] - P[0][0]))
            shift_y = int(round(P[i][1] - P[0][1]))

            cur_mask = roll_with_ignore_multidim(mask, [shift_y, shift_x])
            warped_masks.append(cur_mask)
            
            
        warped_masks = [v[..., None] for v in warped_masks]
            
    warped_masks = np.stack(warped_masks, axis=0) # [f, h, w]
    warped_masks = np.repeat(warped_masks, 3, axis=-1) # [f, h, w, 3]
    
    mask_sum = np.sum(warped_masks, axis=0, keepdims=True)  # [1, H, W, 3]
    mask_sum[mask_sum > 1.0] = 1.0
    mask_sum = mask_sum[0,:,:, 0]
    
    if DEBUG:
        ## visulize warp mask    
        warp_masks_vis = torch.tensor(warped_masks)
        warp_masks_vis = (warp_masks_vis * 255.0).to(torch.uint8)
        torchvision.io.write_video(f'{cur_OUTPUT_PATH}/warped_masks.mp4', warp_masks_vis, fps=10, video_codec='h264', options={'crf': '10'})
        
        # visulize mask
        mask_sum_vis = mask_sum
        mask_sum_vis = (mask_sum_vis * 255.0).astype(np.uint8)
        mask_sum_vis = Image.fromarray(mask_sum_vis)
        
        mask_sum_vis.save(f'{cur_OUTPUT_PATH}/merged_mask.png')
        
    if scale_wise_masks:
        min_area = H * W * area_ratio # cal in downscale
        non_zero_len = mask_sum.sum() 
        
        print(f'non_zero_len: {non_zero_len}, min_area: {min_area}')
        
        if non_zero_len > min_area:
            kernel_sizes = [1, 1, 1, 3]
        elif non_zero_len > min_area * 0.5:
            kernel_sizes = [3, 1, 1, 5]
        else:
            kernel_sizes = [5, 3, 3, 7]
    else:
        kernel_sizes = [1, 1, 1, 1]
        
    mask = torch.from_numpy(mask_sum) # [h, w]
    mask = mask[None, None, ...] # [1, 1, h, w]
    mask = F.interpolate(mask, (height, width), mode='bilinear', align_corners=False) # [1, 1, H, W]
    # mask = mask.repeat(1, num_frames, 1, 1) # [1, f, H, W]
    mask = mask.to(pipeline.dtype).to(device)
    
    ##### Mask End ######
    
    ### Got blending pose features Start ###

    pose_features = []
    for i in range(0, len(cur_pose_features)):
        kernel_size = kernel_sizes[i]
        h, w = cur_pose_features[i].shape[-2:]
        
        if fix_pose_features is None:
            pose_features.append(torch.zeros_like(cur_pose_features[i]))
        else:
            pose_features.append(fix_pose_features[i])
            
        cur_mask = F.interpolate(mask, (h, w), mode='bilinear', align_corners=False)
        cur_mask = dilate_mask_pytorch(cur_mask, kernel_size=kernel_size) # [1, 1, H, W]
        cur_mask = cur_mask.repeat(1, num_frames, 1, 1) # [1, f, H, W]
        
        if DEBUG:
            # visulize mask
            mask_vis = cur_mask[0, 0].cpu().numpy() * 255.0
            mask_vis = Image.fromarray(mask_vis.astype(np.uint8))
            mask_vis.save(f'{cur_OUTPUT_PATH}/mask_k{kernel_size}_scale{i}.png')
            
        cur_mask = cur_mask[None, ...] # [1, 1, f, H, W]
        pose_features[-1] = cur_pose_features[i] * cur_mask + pose_features[-1] * (1 - cur_mask)

    ### Got blending pose features End ###
    
    ##### Warp Noise Start ######
    
    if shared_wapring_latents:
        noise = latents_org[0, 0].data.cpu().numpy().copy() #[14, 4, 40, 72]
        noise = np.transpose(noise, (1, 2, 0)) # [40, 72, 4]

        try:
            warp_noise = Unprojected(noise, depth_down, RTs, H=H, W=W, K=K)
            warp_noise.insert(0, noise)
        except:
            print(f'!!! Noise is too small to warp; mask to bbox')
            
            warp_noise = [noise]
            for i in range(1, num_frames):
                shift_x = int(round(P[i][0] - P[0][0]))
                shift_y = int(round(P[i][1] - P[0][1]))
                
                cur_noise= roll_with_ignore_multidim(noise, [shift_y, shift_x])
                warp_noise.append(cur_noise)
                
            warp_noise = np.stack(warp_noise, axis=0) # [f, h, w, 4]
    
        if DEBUG:
            ## visulize warp noise
            warp_noise_vis = torch.tensor(warp_noise)[..., :3] * torch.tensor(warped_masks)
            warp_noise_vis = (warp_noise_vis - warp_noise_vis.min()) / (warp_noise_vis.max() - warp_noise_vis.min())
            warp_noise_vis = (warp_noise_vis * 255.0).to(torch.uint8)
    
            torchvision.io.write_video(f'{cur_OUTPUT_PATH}/warp_noise.mp4', warp_noise_vis, fps=10, video_codec='h264', options={'crf': '10'})
    
    
        warp_latents = torch.tensor(warp_noise).permute(0, 3, 1, 2).to(latents_org.device).to(latents_org.dtype) # [frame, 4, H, W]
        warp_latents = warp_latents.unsqueeze(0) # [1, frame, 4, H, W]
        
        warped_masks = torch.tensor(warped_masks).permute(0, 3, 1, 2).unsqueeze(0) # [1, frame, 3, H, W]
        mask_extend = torch.concat([warped_masks, warped_masks[:,:,0:1]], dim=2) # [1, frame, 4, H, W]
        mask_extend = mask_extend.to(latents_org.device).to(latents_org.dtype)
        
        warp_latents = warp_latents * mask_extend + latents_org * (1 - mask_extend)
        warp_latents = warp_latents.permute(0, 2, 1, 3, 4)
        random_noise = latents_org.clone().permute(0, 2, 1, 3, 4)
            
        filter_shape = warp_latents.shape

        freq_filter = get_freq_filter(
            filter_shape, 
            device = device, 
            filter_type='butterworth',
            n=4,
            d_s=ds,
            d_t=dt
        )
        
        warp_latents = freq_mix_3d(warp_latents, random_noise, freq_filter)
        warp_latents = warp_latents.permute(0, 2, 1, 3, 4)
        
    else:
        warp_latents = latents_org.clone()
        
    generator.manual_seed(42)

    with torch.no_grad():
        result = pipeline(
            image=condition_image,
            pose_embedding=cur_plucker_embedding,
            height=height,
            width=width,
            num_frames=num_frames,
            num_inference_steps=num_inference_steps,
            min_guidance_scale=min_guidance_scale,
            max_guidance_scale=max_guidance_scale,
            do_image_process=True,
            generator=generator,
            output_type='pt',
            pose_features= pose_features,
            latents = warp_latents
        ).frames[0].cpu() #[f, c, h, w]
        
    
    result = rearrange(result, 'f c h w -> f h w c')
    result = (result * 255.0).to(torch.uint8)

    video_path = tempfile.NamedTemporaryFile(suffix='.mp4').name
    torchvision.io.write_video(video_path, result, fps=10, video_codec='h264', options={'crf': '8'})
    
    return video_path

# -------------- UI definition --------------
with gr.Blocks() as demo:
    # layout definition
    gr.Markdown(title)
    gr.Markdown(authors)
    gr.Markdown(affiliation)
    gr.Markdown(important_link)
    gr.Markdown(description)
    
    
    # with gr.Row():
    #     gr.Markdown("""# <center>Repositioning the Subject within Image </center>""")
    mask = gr.State(value=None) # store mask
    removal_mask = gr.State(value=None) # store removal mask
    selected_points = gr.State([]) # store points
    selected_points_text = gr.Textbox(label="Selected Points", visible=False)
    
    original_image = gr.State(value=None) # store original input image
    masked_original_image = gr.State(value=None) # store masked input image
    mask_logits = gr.State(value=None) # store mask logits
    
    depth = gr.State(value=None) # store depth
    org_depth_image = gr.State(value=None) # store original depth image
    
    camera_pose = gr.State(value=None) # store camera pose
    
    with gr.Column():
        
        outlines = """
        <font size="5"><b>There are total 5 steps to complete the task.</b></font>
        - Step 1: Input an image and Crop it to a suitable size;
        - Step 2: Attain the subject mask;
        - Step 3: Get depth and Draw Trajectory;
        - Step 4: Get camera pose from trajectory or customize it;
        - Step 5: Generate the final video.
        """
        
        gr.Markdown(outlines)
        
        
        with gr.Row():
            with gr.Column():
                # Step 1: Input Image
                step1_dec = """
                    <font size="4"><b>Step 1: Input Image</b></font>
                    - Select the region using a <mark>bounding box</mark>, aiming for a ratio close to </mark>320:576</mark> (height:width).
                    - All provided images in `Examples` are in 320 x 576 resolution. Simply press `Process` to proceed.
                    """
                step1 = gr.Markdown(step1_dec)
                raw_input = ImagePrompter(type="pil", label="Raw Image", show_label=True, interactive=True)
                # left_up_point = gr.Textbox(value = "-1 -1", label="Left Up Point", interactive=True)
                process_button = gr.Button("Process")
                
            with gr.Column():
                # Step 2: Get Subject Mask
                step2_dec = """
                    <font size="4"><b>Step 2: Get Subject Mask</b></font>
                    - Use the <mark>bounding boxes</mark> or <mark>paints</mark> to select the subject.
                    - Press `Segment Subject` to get the mask. <mark>Can be refined iteratively by updating points<mark>.
                    """
                step2 = gr.Markdown(step2_dec)
                canvas = ImagePrompter(type="pil", label="Input Image", show_label=True, interactive=True) # for mask painting

                select_button = gr.Button("Segment Subject")
                
        with gr.Row():
            with gr.Column():
                mask_dec = """
                    <font size="4"><b>Mask Result</b></font>
                    - Just for visualization purpose. No need to interact.
                """
                mask_vis = gr.Markdown(mask_dec)
                mask_output = gr.Image(type="pil", label="Mask", show_label=True, interactive=False)
            with gr.Column():
                # Step 3: Get Depth and Draw Trajectory
                step3_dec = """
                    <font size="4"><b>Step 3: Get Depth and Draw Trajectory</b></font>
                    - Press `Get Depth` to get the depth image.
                    - Draw the trajectory by selecting points on the depth image. <mark>No more than 14 points</mark>.
                    - Press `Undo point` to remove all points.
                """
                step3 = gr.Markdown(step3_dec)
                depth_image = gr.Image(type="pil", label="Depth Image", show_label=True, interactive=False)
                with gr.Row():
                    depth_button = gr.Button("Get Depth")
                    undo_button = gr.Button("Undo point")
                    
        with gr.Row():
            with gr.Column():
                # Step 4: Trajectory to Camera Pose or Get Camera Pose
                step4_dec = """
                    <font size="4"><b>Step 4: Get camera pose from trajectory or customize it</b></font>
                    - Option 1: Transform the 2D trajectory to camera poses with depth. <mark>`Rescale` is used for depth alignment. Larger value can speed up the object motion.</mark>
                    - Option 2: Rotate the camera with a specific `Angle`.
                    - Option 3: Rotate the camera clockwise or counterclockwise with a specific `Angle`.
                    - Option 4: Translate the camera with `Tx` (<mark>Pan Left/Right</mark>), `Ty` (<mark>Pan Up/Down</mark>), `Tz` (<mark>Zoom In/Out</mark>) and `Speed`.
                """
                step4 = gr.Markdown(step4_dec)
                camera_pose_vis = gr.Plot(None, label='Camera Pose')
                with gr.Row():
                    with gr.Column():
                        speed = gr.Slider(minimum=0.1, maximum=10, step=0.1, value=1.0, label="Speed", interactive=True)
                        rescale = gr.Slider(minimum=0.0, maximum=10, step=0.1, value=1.0, label="Rescale", interactive=True)
                        # traj2pose_button = gr.Button("Option1: Trajectory to Camera Pose")
                        
                        angle = gr.Slider(minimum=-360, maximum=360, step=1, value=60, label="Angle", interactive=True)
                        # rotation_button = gr.Button("Option2: Rotate")
                        # clockwise_button = gr.Button("Option3: Clockwise")
                    with gr.Column():
                        
                        Tx = gr.Slider(minimum=-1, maximum=1, step=1, value=0, label="Tx", interactive=True)
                        Ty = gr.Slider(minimum=-1, maximum=1, step=1, value=0, label="Ty", interactive=True)
                        Tz = gr.Slider(minimum=-1, maximum=1, step=1, value=0, label="Tz", interactive=True)
                        # translation_button = gr.Button("Option4: Translate")
                with gr.Row():
                    camera_option = gr.Radio(choices = CAMERA_MODE, label='Camera Options', value=CAMERA_MODE[0], interactive=True)
                with gr.Row():
                    get_camera_pose_button = gr.Button("Get Camera Pose")
                        
            with gr.Column():
                # Step 5: Get the final generated video
                step5_dec = """
                    <font size="4"><b>Step 5: Get the final generated video</b></font>
                    - 3 modes for background: <mark>Fixed</mark>, <mark>Reverse</mark>, <mark>Free</mark>.
                    - Enable <mark>Scale-wise Masks</mark> for better object control.
                    - Option to enable <mark>Shared Warping Latents</mark> and set <mark>stop frequency</mark> for spatial (`ds`) and temporal (`dt`) dimensions. Larger stop frequency will lead to artifacts.
                """
                step5 = gr.Markdown(step5_dec)
                generated_video = gr.Video(None, label='Generated Video')
                
                with gr.Row():
                    seed = gr.Textbox(value = "42", label="Seed", interactive=True)
                    # num_inference_steps = gr.Slider(minimum=1, maximum=100, step=1, value=25, label="Number of Inference Steps", interactive=True)
                    bg_mode = gr.Radio(choices = ["Fixed", "Reverse", "Free"], label="Background Mode", value="Fixed", interactive=True)
                # swl_mode = gr.Radio(choices = ["Enable SWL", "Disable SWL"], label="Shared Warping Latent", value="Disable SWL", interactive=True)
                scale_wise_masks = gr.Checkbox(label="Enable Scale-wise Masks", interactive=True, value=True)
                with gr.Row():
                    with gr.Column():
                        shared_wapring_latents = gr.Checkbox(label="Enable Shared Warping Latents", interactive=True)
                    with gr.Column():
                        ds = gr.Slider(minimum=0.0, maximum=1, step=0.1, value=0.5, label="ds", interactive=True)
                        dt = gr.Slider(minimum=0.0, maximum=1, step=0.1, value=0.5, label="dt", interactive=True)
                
                generated_button = gr.Button("Generate")

                

    # # event definition
    process_button.click(
        fn = process_image,
        inputs = [raw_input],
        outputs = [original_image, canvas]
    )
    
    select_button.click(
        segment,
        [canvas, original_image, mask_logits],
        [mask, mask_output, masked_original_image, mask_logits]
    )
    
    depth_button.click(
        get_depth,
        [original_image, selected_points],
        [depth, depth_image, org_depth_image]
    )
    
    depth_image.select(
        get_points,
        [depth_image, selected_points],
        [depth_image, selected_points],
    )
    undo_button.click(
        undo_points,
        [org_depth_image],
        [depth_image, selected_points]
    )
    
    get_camera_pose_button.click(
        get_camera_pose(CAMERA_MODE),
        [camera_option, selected_points, depth, mask, rescale, angle, Tx, Ty, Tz, speed],
        [camera_pose, camera_pose_vis, rescale]
    )
    
    generated_button.click(
        run_objctrl_2_5d,
        [
         original_image,
         mask,
         depth,
         camera_pose,
         bg_mode,
         shared_wapring_latents,
         scale_wise_masks,
         rescale,
         seed,
         ds,
         dt,
        #  num_inference_steps
         ],
        [generated_video],
    )

    gr.Examples(
        examples=examples,
        inputs=[
            raw_input,
            rescale,
            speed,
            angle,
            Tx,
            Ty,
            Tz,
            camera_option,
            bg_mode,
            shared_wapring_latents,
            scale_wise_masks,
            ds,
            dt,
            seed,
            selected_points_text  # selected_points
        ],
        outputs=[generated_video], 
        examples_per_page=10
    )
    
    selected_points_text.change(
        sync_points,
        inputs=[selected_points_text],
        outputs=[selected_points]
    )


    gr.Markdown(article)


demo.queue().launch(share=True)