Spaces:
Runtime error
Runtime error
allow manual keypoints at edit hands; put fixed hand to original image
Browse files- .gitignore +6 -0
- README.md +1 -1
- app.py +563 -254
- brown_logo.png +3 -0
- meta_logo.png +3 -0
- sbatch/sbatch_demo.sh +38 -0
- vqvae.py +4 -1
.gitignore
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
settings.json
|
2 |
+
sbatch/err/
|
3 |
+
sbatch/out/
|
4 |
+
__pycache__/
|
5 |
+
diffusion/__pycache__/
|
6 |
+
*.pyc
|
README.md
CHANGED
@@ -4,7 +4,7 @@ emoji: ✋
|
|
4 |
colorFrom: gray
|
5 |
colorTo: purple
|
6 |
sdk: gradio
|
7 |
-
sdk_version: 4.
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
short_description: FoundHand
|
|
|
4 |
colorFrom: gray
|
5 |
colorTo: purple
|
6 |
sdk: gradio
|
7 |
+
sdk_version: 4.40.1
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
short_description: FoundHand
|
app.py
CHANGED
@@ -20,7 +20,10 @@ from copy import deepcopy
|
|
20 |
from typing import Optional
|
21 |
import requests
|
22 |
from huggingface_hub import hf_hub_download
|
23 |
-
|
|
|
|
|
|
|
24 |
|
25 |
MAX_N = 6
|
26 |
FIX_MAX_N = 6
|
@@ -29,6 +32,12 @@ placeholder = cv2.cvtColor(cv2.imread("placeholder.png"), cv2.COLOR_BGR2RGB)
|
|
29 |
NEW_MODEL = True
|
30 |
MODEL_EPOCH = 6
|
31 |
REF_POSE_MASK = True
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
def set_seed(seed):
|
34 |
seed = int(seed)
|
@@ -112,7 +121,7 @@ def visualize_hand(all_joints, img, side=["right", "left"], n_avail_joints=21):
|
|
112 |
# Convert BytesIO object to numpy array
|
113 |
buf.seek(0)
|
114 |
img_pil = Image.open(buf)
|
115 |
-
img_pil = img_pil.resize((
|
116 |
numpy_img = np.array(img_pil)
|
117 |
|
118 |
return numpy_img
|
@@ -232,31 +241,9 @@ if NEW_MODEL:
|
|
232 |
print(f"encoder after eval() max: {max([p.max() for p in autoencoder.encoder.parameters()])}")
|
233 |
print(f"autoencoder encoder after eval() dtype: {next(autoencoder.encoder.parameters()).dtype}")
|
234 |
assert len(missing_keys) == 0
|
235 |
-
# else:
|
236 |
-
# opts = HandDiffOpts()
|
237 |
-
# model_path = './finetune_epoch=5-step=130000.ckpt'
|
238 |
-
# sd_path = './sd-v1-4.ckpt'
|
239 |
-
# print('Load diffusion model...')
|
240 |
-
# diffusion = create_diffusion(str(opts.test_sampling_steps))
|
241 |
-
# model = vit.DiT_XL_2(
|
242 |
-
# input_size=opts.latent_size[0],
|
243 |
-
# latent_dim=opts.latent_dim,
|
244 |
-
# in_channels=opts.latent_dim+opts.n_keypoints+opts.n_mask,
|
245 |
-
# learn_sigma=True,
|
246 |
-
# ).to(device)
|
247 |
-
# ckpt_state_dict = torch.load(model_path)['state_dict']
|
248 |
-
# dit_state_dict = {remove_prefix(k, 'diffusion_backbone.'): v for k, v in ckpt_state_dict.items() if k.startswith('diffusion_backbone')}
|
249 |
-
# vae_state_dict = {remove_prefix(k, 'autoencoder.'): v for k, v in ckpt_state_dict.items() if k.startswith('autoencoder')}
|
250 |
-
# missing_keys, extra_keys = model.load_state_dict(dit_state_dict, strict=False)
|
251 |
-
# model.eval()
|
252 |
-
# assert len(missing_keys) == 0 and len(extra_keys) == 0
|
253 |
-
# autoencoder = vqvae.create_model(3, 3, opts.latent_dim).eval().requires_grad_(False).to(device)
|
254 |
-
# missing_keys, extra_keys = autoencoder.load_state_dict(vae_state_dict, strict=False)
|
255 |
-
# autoencoder.eval()
|
256 |
-
# assert len(missing_keys) == 0 and len(extra_keys) == 0
|
257 |
-
sam_path = hf_hub_download(repo_id="Chaerin5/FoundHand-weights", filename="sam_vit_h_4b8939.pth", token=token)
|
258 |
-
sam_predictor = init_sam(ckpt_path=sam_path, device='cpu')
|
259 |
|
|
|
|
|
260 |
|
261 |
print("Mediapipe hand detector and SAM ready...")
|
262 |
mp_hands = mp.solutions.hands
|
@@ -266,17 +253,12 @@ hands = mp_hands.Hands(
|
|
266 |
min_detection_confidence=0.1,
|
267 |
)
|
268 |
|
269 |
-
def
|
270 |
if ref is None:
|
271 |
return (
|
272 |
-
None,
|
273 |
-
None,
|
274 |
-
None,
|
275 |
-
None,
|
276 |
-
None,
|
277 |
)
|
278 |
-
missing_keys, extra_keys = autoencoder.load_state_dict(vae_state_dict, strict=False)
|
279 |
-
|
280 |
img = ref["composite"][..., :3]
|
281 |
img = cv2.resize(img, opts.image_size, interpolation=cv2.INTER_AREA)
|
282 |
keypts = np.zeros((42, 2))
|
@@ -307,6 +289,7 @@ def get_ref_anno(img, keypts):
|
|
307 |
if keypts is None:
|
308 |
no_hands = cv2.resize(np.array(Image.open("no_hands.png"))[..., :3], (LENGTH, LENGTH))
|
309 |
return None, no_hands, None
|
|
|
310 |
if isinstance(keypts, list):
|
311 |
if len(keypts[0]) == 0:
|
312 |
keypts[0] = np.zeros((21, 2))
|
@@ -315,7 +298,6 @@ def get_ref_anno(img, keypts):
|
|
315 |
else:
|
316 |
gr.Info("Number of right hand keypoints should be either 0 or 21.")
|
317 |
return None, None, None
|
318 |
-
|
319 |
if len(keypts[1]) == 0:
|
320 |
keypts[1] = np.zeros((21, 2))
|
321 |
elif len(keypts[1]) == 21:
|
@@ -323,7 +305,6 @@ def get_ref_anno(img, keypts):
|
|
323 |
else:
|
324 |
gr.Info("Number of left hand keypoints should be either 0 or 21.")
|
325 |
return None, None, None
|
326 |
-
|
327 |
keypts = np.concatenate(keypts, axis=0)
|
328 |
if REF_POSE_MASK:
|
329 |
sam_predictor.set_image(img)
|
@@ -362,7 +343,7 @@ def get_ref_anno(img, keypts):
|
|
362 |
Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
|
363 |
]
|
364 |
)
|
365 |
-
image = image_transform(img)
|
366 |
kpts_valid = check_keypoints_validity(keypts, target_size)
|
367 |
heatmaps = torch.tensor(
|
368 |
keypoint_heatmap(
|
@@ -370,7 +351,7 @@ def get_ref_anno(img, keypts):
|
|
370 |
)
|
371 |
* kpts_valid[:, None, None],
|
372 |
dtype=torch.float,
|
373 |
-
|
374 |
)[None, ...]
|
375 |
mask = torch.tensor(
|
376 |
cv2.resize(
|
@@ -379,7 +360,7 @@ def get_ref_anno(img, keypts):
|
|
379 |
interpolation=cv2.INTER_NEAREST,
|
380 |
),
|
381 |
dtype=torch.float,
|
382 |
-
|
383 |
).unsqueeze(0)[None, ...]
|
384 |
return image[None, ...], heatmaps, mask
|
385 |
|
@@ -388,7 +369,7 @@ def get_ref_anno(img, keypts):
|
|
388 |
img,
|
389 |
keypts,
|
390 |
hand_mask,
|
391 |
-
device=
|
392 |
target_size=opts.image_size,
|
393 |
latent_size=opts.latent_size,
|
394 |
)
|
@@ -409,62 +390,49 @@ def get_ref_anno(img, keypts):
|
|
409 |
|
410 |
return img, ref_pose, ref_cond
|
411 |
|
412 |
-
def get_target_anno(
|
413 |
-
if
|
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 |
-
detected[1] = 1
|
440 |
-
for i, landmark in enumerate(hand_landmarks.landmark):
|
441 |
-
target_keypts[start_idx + i] = [
|
442 |
-
landmark.x * opts.image_size[1],
|
443 |
-
landmark.y * opts.image_size[0],
|
444 |
-
]
|
445 |
-
|
446 |
-
target_pose = visualize_hand(target_keypts, pose_img)
|
447 |
-
kpts_valid = check_keypoints_validity(target_keypts, opts.image_size)
|
448 |
-
target_heatmaps = torch.tensor(
|
449 |
-
keypoint_heatmap(
|
450 |
-
scale_keypoint(target_keypts, opts.image_size, opts.latent_size),
|
451 |
-
opts.latent_size,
|
452 |
-
var=1.0,
|
453 |
-
)
|
454 |
-
* kpts_valid[:, None, None],
|
455 |
-
dtype=torch.float,
|
456 |
-
# device=device,
|
457 |
-
)[None, ...]
|
458 |
-
target_cond = torch.cat(
|
459 |
-
[target_heatmaps, torch.zeros_like(target_heatmaps)[:, :1]], 1
|
460 |
)
|
461 |
-
|
462 |
-
|
|
|
|
|
|
|
|
|
|
|
463 |
|
464 |
-
return
|
465 |
|
466 |
|
467 |
def get_mask_inpaint(ref):
|
|
|
|
|
|
|
468 |
inpaint_mask = np.array(ref["layers"][0])[..., -1]
|
469 |
inpaint_mask = cv2.resize(
|
470 |
inpaint_mask, opts.image_size, interpolation=cv2.INTER_AREA
|
@@ -473,12 +441,12 @@ def get_mask_inpaint(ref):
|
|
473 |
return inpaint_mask
|
474 |
|
475 |
|
476 |
-
def visualize_ref(crop,
|
477 |
-
if
|
478 |
return None
|
479 |
inpainted = brush["layers"][0][..., -1]
|
480 |
-
img =
|
481 |
-
img = cv2.resize(img, inpainted.shape[::-1], interpolation=cv2.INTER_AREA)
|
482 |
mask = inpainted < 128
|
483 |
# img = img.astype(np.int32)
|
484 |
# img[mask, :] = img[mask, :] - 50
|
@@ -539,7 +507,39 @@ def reset_kps(img, keypoints, side: Literal["right", "left"]):
|
|
539 |
keypoints[1] = []
|
540 |
return img, keypoints
|
541 |
|
542 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
543 |
def sample_diff(ref_cond, target_cond, target_keypts, num_gen, seed, cfg):
|
544 |
set_seed(seed)
|
545 |
z = torch.randn(
|
@@ -586,14 +586,17 @@ def sample_diff(ref_cond, target_cond, target_keypts, num_gen, seed, cfg):
|
|
586 |
print(f"results[0].max(): {results[0].max()}")
|
587 |
return results, results_pose
|
588 |
|
589 |
-
@
|
590 |
-
def ready_sample(
|
591 |
-
img = cv2.resize(img_ori[..., :3], opts.image_size, interpolation=cv2.INTER_AREA)
|
|
|
592 |
sam_predictor.set_image(img)
|
593 |
if len(keypts[0]) == 0:
|
594 |
keypts[0] = np.zeros((21, 2))
|
595 |
elif len(keypts[0]) == 21:
|
596 |
keypts[0] = np.array(keypts[0], dtype=np.float32)
|
|
|
|
|
597 |
else:
|
598 |
gr.Info("Number of right hand keypoints should be either 0 or 21.")
|
599 |
return None, None
|
@@ -602,12 +605,14 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
602 |
keypts[1] = np.zeros((21, 2))
|
603 |
elif len(keypts[1]) == 21:
|
604 |
keypts[1] = np.array(keypts[1], dtype=np.float32)
|
|
|
|
|
605 |
else:
|
606 |
gr.Info("Number of left hand keypoints should be either 0 or 21.")
|
607 |
return None, None
|
608 |
|
609 |
keypts = np.concatenate(keypts, axis=0)
|
610 |
-
keypts = scale_keypoint(keypts, (
|
611 |
|
612 |
box_shift_ratio = 0.5
|
613 |
box_size_factor = 1.2
|
@@ -643,7 +648,7 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
643 |
inpaint_mask, dsize=opts.latent_size, interpolation=cv2.INTER_NEAREST
|
644 |
),
|
645 |
dtype=torch.float,
|
646 |
-
|
647 |
).unsqueeze(0)[None, ...]
|
648 |
|
649 |
def make_ref_cond(
|
@@ -661,7 +666,7 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
661 |
Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
|
662 |
]
|
663 |
)
|
664 |
-
image = image_transform(img)
|
665 |
kpts_valid = check_keypoints_validity(keypts, target_size)
|
666 |
heatmaps = torch.tensor(
|
667 |
keypoint_heatmap(
|
@@ -669,7 +674,7 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
669 |
)
|
670 |
* kpts_valid[:, None, None],
|
671 |
dtype=torch.float,
|
672 |
-
|
673 |
)[None, ...]
|
674 |
mask = torch.tensor(
|
675 |
cv2.resize(
|
@@ -678,7 +683,7 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
678 |
interpolation=cv2.INTER_NEAREST,
|
679 |
),
|
680 |
dtype=torch.float,
|
681 |
-
|
682 |
).unsqueeze(0)[None, ...]
|
683 |
return image[None, ...], heatmaps, mask
|
684 |
|
@@ -686,7 +691,7 @@ def ready_sample(img_ori, inpaint_mask, keypts):
|
|
686 |
img,
|
687 |
keypts,
|
688 |
hand_mask * (1 - inpaint_mask),
|
689 |
-
device=
|
690 |
target_size=opts.image_size,
|
691 |
latent_size=opts.latent_size,
|
692 |
)
|
@@ -726,13 +731,15 @@ def switch_mask_size(radio):
|
|
726 |
out = (gr.update(visible=True), gr.update(visible=False))
|
727 |
return out
|
728 |
|
729 |
-
@
|
730 |
def sample_inpaint(
|
731 |
ref_cond,
|
732 |
target_cond,
|
733 |
latent,
|
734 |
inpaint_latent_mask,
|
735 |
keypts,
|
|
|
|
|
736 |
num_gen,
|
737 |
seed,
|
738 |
cfg,
|
@@ -778,39 +785,76 @@ def sample_inpaint(
|
|
778 |
# visualize
|
779 |
results = []
|
780 |
results_pose = []
|
|
|
781 |
for i in range(FIX_MAX_N):
|
782 |
if i < num_gen:
|
783 |
-
|
784 |
-
|
|
|
|
|
|
|
|
|
|
|
785 |
else:
|
786 |
results.append(placeholder)
|
787 |
results_pose.append(placeholder)
|
788 |
-
|
|
|
789 |
|
790 |
|
791 |
def flip_hand(
|
792 |
-
img,
|
793 |
-
manual_kp_right
|
|
|
|
|
794 |
):
|
795 |
if cond is None: # clear clicked
|
796 |
-
return
|
797 |
img["composite"] = img["composite"][:, ::-1, :]
|
798 |
img["background"] = img["background"][:, ::-1, :]
|
799 |
img["layers"] = [layer[:, ::-1, :] for layer in img["layers"]]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
800 |
pose_img = pose_img[:, ::-1, :]
|
|
|
|
|
|
|
|
|
|
|
|
|
801 |
cond = cond.flip(-1)
|
802 |
-
if
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
803 |
if keypts[:21, :].sum() != 0:
|
804 |
keypts[:21, 0] = opts.image_size[1] - keypts[:21, 0]
|
805 |
-
# keypts[:21, 1] = opts.image_size[0] - keypts[:21, 1]
|
806 |
if keypts[21:, :].sum() != 0:
|
807 |
keypts[21:, 0] = opts.image_size[1] - keypts[21:, 0]
|
808 |
-
|
809 |
-
|
810 |
-
|
811 |
-
|
812 |
-
|
813 |
-
|
|
|
|
|
|
|
|
|
|
|
814 |
|
815 |
|
816 |
def resize_to_full(img):
|
@@ -823,26 +867,30 @@ def resize_to_full(img):
|
|
823 |
def clear_all():
|
824 |
return (
|
825 |
None,
|
|
|
826 |
None,
|
827 |
None,
|
828 |
None,
|
829 |
None,
|
830 |
-
False,
|
831 |
None,
|
832 |
None,
|
833 |
False,
|
834 |
None,
|
835 |
None,
|
|
|
|
|
|
|
|
|
836 |
None,
|
837 |
None,
|
838 |
None,
|
|
|
839 |
None,
|
840 |
None,
|
841 |
1,
|
842 |
42,
|
843 |
3.0,
|
844 |
gr.update(interactive=False),
|
845 |
-
[]
|
846 |
)
|
847 |
|
848 |
|
@@ -851,6 +899,9 @@ def fix_clear_all():
|
|
851 |
None,
|
852 |
None,
|
853 |
None,
|
|
|
|
|
|
|
854 |
None,
|
855 |
None,
|
856 |
None,
|
@@ -876,14 +927,14 @@ def fix_clear_all():
|
|
876 |
def enable_component(image1, image2):
|
877 |
if image1 is None or image2 is None:
|
878 |
return gr.update(interactive=False)
|
879 |
-
if "background" in image1 and "layers" in image1 and "composite" in image1:
|
880 |
if (
|
881 |
image1["background"].sum() == 0
|
882 |
and (sum([im.sum() for im in image1["layers"]]) == 0)
|
883 |
and image1["composite"].sum() == 0
|
884 |
):
|
885 |
return gr.update(interactive=False)
|
886 |
-
if "background" in image2 and "layers" in image2 and "composite" in image2:
|
887 |
if (
|
888 |
image2["background"].sum() == 0
|
889 |
and (sum([im.sum() for im in image2["layers"]]) == 0)
|
@@ -940,6 +991,18 @@ def set_visible(checkbox, kpts, img_clean, img_pose_right, img_pose_left, done=N
|
|
940 |
|
941 |
def set_unvisible():
|
942 |
return (
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
943 |
gr.update(visible=False),
|
944 |
gr.update(visible=False),
|
945 |
gr.update(visible=False),
|
@@ -954,6 +1017,18 @@ def set_unvisible():
|
|
954 |
gr.update(visible=False)
|
955 |
)
|
956 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
957 |
def set_no_hands(decider, component):
|
958 |
if decider is None:
|
959 |
no_hands = cv2.resize(np.array(Image.open("no_hands.png"))[..., :3], (LENGTH, LENGTH))
|
@@ -975,19 +1050,6 @@ def unvisible_component(decider, component):
|
|
975 |
update_component = gr.update(visible=True)
|
976 |
return update_component
|
977 |
|
978 |
-
# def make_change(decider, state):
|
979 |
-
# '''
|
980 |
-
# if decider is not None, change the state's value. True/False does not matter.
|
981 |
-
# '''
|
982 |
-
# if decider is not None:
|
983 |
-
# if state:
|
984 |
-
# state = False
|
985 |
-
# else:
|
986 |
-
# state = True
|
987 |
-
# return state
|
988 |
-
# else:
|
989 |
-
# return state
|
990 |
-
|
991 |
LENGTH = 480
|
992 |
|
993 |
example_ref_imgs = [
|
@@ -1083,7 +1145,7 @@ fix_example_imgs = [
|
|
1083 |
# ["bad_hands/4.jpg"], # "bad_hands/4_mask.jpg"],
|
1084 |
["bad_hands/5.jpg"], # "bad_hands/5_mask.jpg"],
|
1085 |
["bad_hands/6.jpg"], # "bad_hands/6_mask.jpg"],
|
1086 |
-
["bad_hands/7.jpg"], # "bad_hands/7_mask.jpg"],
|
1087 |
# ["bad_hands/8.jpg"], # "bad_hands/8_mask.jpg"],
|
1088 |
# ["bad_hands/9.jpg"], # "bad_hands/9_mask.jpg"],
|
1089 |
# ["bad_hands/10.jpg"], # "bad_hands/10_mask.jpg"],
|
@@ -1137,20 +1199,32 @@ _CITE_ = r"""
|
|
1137 |
with gr.Blocks(css=custom_css, theme="soft") as demo:
|
1138 |
gr.Markdown(_HEADER_)
|
1139 |
with gr.Tab("Edit Hand Poses"):
|
|
|
|
|
|
|
1140 |
ref_img = gr.State(value=None)
|
1141 |
ref_im_raw = gr.State(value=None)
|
1142 |
ref_kp_raw = gr.State(value=0)
|
1143 |
ref_kp_got = gr.State(value=None)
|
1144 |
-
dump = gr.State(value=None)
|
1145 |
-
ref_cond = gr.State(value=None)
|
1146 |
ref_manual_cond = gr.State(value=None)
|
1147 |
ref_auto_cond = gr.State(value=None)
|
1148 |
-
|
|
|
|
|
1149 |
target_img = gr.State(value=None)
|
1150 |
-
|
|
|
|
|
|
|
|
|
1151 |
target_keypts = gr.State(value=None)
|
1152 |
-
|
|
|
|
|
|
|
|
|
1153 |
with gr.Row():
|
|
|
1154 |
with gr.Column():
|
1155 |
gr.Markdown(
|
1156 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">1. Upload a hand image to edit 📥</p>"""
|
@@ -1270,6 +1344,8 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1270 |
ref_flip = gr.Checkbox(
|
1271 |
value=False, label="Flip Handedness (Reference)", interactive=False
|
1272 |
)
|
|
|
|
|
1273 |
with gr.Column():
|
1274 |
gr.Markdown(
|
1275 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">2. Upload a hand image for target hand pose 📥</p>"""
|
@@ -1294,20 +1370,105 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1294 |
target_finish_crop = gr.Button(
|
1295 |
value="Finish Cropping", interactive=False
|
1296 |
)
|
1297 |
-
|
1298 |
-
|
1299 |
-
|
1300 |
-
|
1301 |
-
|
1302 |
-
|
1303 |
-
|
1304 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1305 |
gr.Markdown(
|
1306 |
"""<p style="text-align: center;">③ Optionally flip the hand</p>"""
|
1307 |
)
|
1308 |
target_flip = gr.Checkbox(
|
1309 |
value=False, label="Flip Handedness (Target)", interactive=False
|
1310 |
)
|
|
|
|
|
1311 |
with gr.Column():
|
1312 |
gr.Markdown(
|
1313 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">3. Press "Run" to get the edited results 🎯</p>"""
|
@@ -1371,10 +1532,18 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1371 |
interactive=True,
|
1372 |
)
|
1373 |
|
|
|
1374 |
ref.change(enable_component, [ref, ref], ref_finish_crop)
|
1375 |
-
ref_finish_crop.click(
|
1376 |
ref_kp_raw.change(lambda x: x, ref_im_raw, ref_manual_kp_right)
|
1377 |
ref_kp_raw.change(lambda x: x, ref_im_raw, ref_manual_kp_left)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1378 |
ref_manual_checkbox.select(
|
1379 |
set_visible,
|
1380 |
[ref_manual_checkbox, ref_kp_got, ref_im_raw, ref_manual_kp_right, ref_manual_kp_left, ref_manual_done],
|
@@ -1412,38 +1581,94 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1412 |
ref_manual_reset_left.click(
|
1413 |
reset_kps, [ref_im_raw, ref_kp_got, gr.State("left")], [ref_manual_kp_left, ref_kp_got]
|
1414 |
)
|
|
|
|
|
1415 |
ref_manual_done.click(get_ref_anno, [ref_im_raw, ref_kp_got], [ref_img, ref_manual_pose, ref_manual_cond])
|
1416 |
-
ref_manual_cond.change(lambda x: x, ref_manual_cond, ref_cond)
|
1417 |
-
ref_use_manual.click(lambda x: x, ref_manual_cond, ref_cond)
|
1418 |
-
# ref_use_manual.click(lambda x: gr.Info("Manual hand keypoints will be used for 'Reference'", duration=3))
|
1419 |
-
ref_manual_done.click(visible_component, [ref_manual_pose, ref_manual_pose], ref_manual_pose)
|
1420 |
-
ref_manual_done.click(visible_component, [ref_use_manual, ref_use_manual], ref_use_manual)
|
1421 |
ref_manual_pose.change(enable_component, [ref_manual_pose, ref_manual_pose], ref_manual_done)
|
1422 |
-
ref_kp_raw.change(get_ref_anno, [ref_im_raw, ref_kp_raw], [ref_img, ref_pose, ref_auto_cond])
|
1423 |
-
ref_auto_cond.change(lambda x: x, ref_auto_cond, ref_cond)
|
1424 |
-
ref_use_auto.click(lambda x: x, ref_auto_cond, ref_cond)
|
1425 |
-
# ref_use_auto.click(lambda x: gr.Info("Automatic hand keypoints will be used for 'Reference'", duration=3))
|
1426 |
-
ref_pose.change(enable_component, [ref_kp_raw, ref_pose], ref_use_auto)
|
1427 |
-
ref_pose.change(enable_component, [ref_img, ref_pose], ref_flip)
|
1428 |
ref_manual_pose.change(enable_component, [ref_img, ref_manual_pose], ref_flip)
|
|
|
|
|
|
|
|
|
1429 |
ref_flip.select(
|
1430 |
-
flip_hand,
|
|
|
|
|
1431 |
)
|
|
|
|
|
1432 |
target.change(enable_component, [target, target], target_finish_crop)
|
1433 |
-
target_finish_crop.click(
|
1434 |
-
|
1435 |
-
|
1436 |
-
|
1437 |
-
)
|
1438 |
target_pose.change(enable_component, [target_img, target_pose], target_flip)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1439 |
target_flip.select(
|
1440 |
flip_hand,
|
1441 |
-
[target, target_pose, target_cond, target_keypts],
|
1442 |
-
[target, target_pose, target_cond, target_keypts],
|
1443 |
)
|
1444 |
-
|
1445 |
-
|
1446 |
-
|
|
|
|
|
|
|
1447 |
run.click(
|
1448 |
sample_diff,
|
1449 |
[ref_cond, target_cond, target_keypts, n_generation, seed, cfg],
|
@@ -1454,34 +1679,40 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1454 |
[],
|
1455 |
[
|
1456 |
ref,
|
|
|
1457 |
ref_manual_kp_right,
|
1458 |
ref_manual_kp_left,
|
|
|
1459 |
ref_pose,
|
1460 |
ref_manual_pose,
|
|
|
1461 |
ref_flip,
|
1462 |
target,
|
|
|
|
|
|
|
|
|
|
|
1463 |
target_pose,
|
|
|
|
|
1464 |
target_flip,
|
1465 |
results,
|
1466 |
results_pose,
|
1467 |
-
ref_img,
|
1468 |
-
ref_cond,
|
1469 |
-
target_img,
|
1470 |
-
target_cond,
|
1471 |
-
target_keypts,
|
1472 |
n_generation,
|
1473 |
seed,
|
1474 |
cfg,
|
1475 |
ref_kp_raw,
|
1476 |
-
ref_manual_checkbox
|
1477 |
],
|
1478 |
)
|
1479 |
clear.click(
|
1480 |
set_unvisible,
|
1481 |
[],
|
1482 |
[
|
1483 |
-
ref_manual_kp_r_info,
|
1484 |
ref_manual_kp_l_info,
|
|
|
|
|
|
|
1485 |
ref_manual_undo_left,
|
1486 |
ref_manual_undo_right,
|
1487 |
ref_manual_reset_left,
|
@@ -1490,14 +1721,25 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1490 |
ref_manual_done_info,
|
1491 |
ref_manual_pose,
|
1492 |
ref_use_manual,
|
1493 |
-
|
1494 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1495 |
]
|
1496 |
)
|
1497 |
|
1498 |
with gr.Tab("Fix Hands"):
|
1499 |
fix_inpaint_mask = gr.State(value=None)
|
1500 |
fix_original = gr.State(value=None)
|
|
|
1501 |
fix_img = gr.State(value=None)
|
1502 |
fix_kpts = gr.State(value=None)
|
1503 |
fix_kpts_np = gr.State(value=None)
|
@@ -1506,37 +1748,62 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1506 |
fix_latent = gr.State(value=None)
|
1507 |
fix_inpaint_latent = gr.State(value=None)
|
1508 |
with gr.Row():
|
|
|
1509 |
with gr.Column():
|
1510 |
gr.Markdown(
|
1511 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">1. Upload a malformed hand image to fix 📥</p>"""
|
1512 |
)
|
1513 |
gr.Markdown(
|
1514 |
-
"""<p style="text-align: center;">① Optionally crop the image around the hand</p>"""
|
1515 |
)
|
1516 |
-
fix_crop = gr.ImageEditor(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1517 |
type="numpy",
|
1518 |
sources=["upload", "webcam", "clipboard"],
|
1519 |
-
label="Image
|
1520 |
show_label=True,
|
1521 |
height=LENGTH,
|
1522 |
width=LENGTH,
|
1523 |
-
|
1524 |
-
|
1525 |
-
|
1526 |
-
|
1527 |
-
|
1528 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1529 |
fix_example = gr.Examples(
|
1530 |
fix_example_imgs,
|
1531 |
inputs=[fix_crop],
|
1532 |
examples_per_page=20,
|
1533 |
)
|
1534 |
gr.Markdown(
|
1535 |
-
"""<p style="text-align: center;">② Brush area (e.g., wrong finger) that needs to be fixed.
|
1536 |
)
|
1537 |
fix_ref = gr.ImageEditor(
|
1538 |
type="numpy",
|
1539 |
-
label="Image
|
1540 |
sources=(),
|
1541 |
show_label=True,
|
1542 |
height=LENGTH,
|
@@ -1550,9 +1817,14 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1550 |
container=False,
|
1551 |
interactive=False,
|
1552 |
)
|
|
|
|
|
|
|
1553 |
fix_finish_crop = gr.Button(
|
1554 |
value="Finish Croping & Brushing", interactive=False
|
1555 |
)
|
|
|
|
|
1556 |
with gr.Column():
|
1557 |
gr.Markdown(
|
1558 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">2. Click on hand to get target hand pose</p>"""
|
@@ -1565,13 +1837,14 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1565 |
show_label=False,
|
1566 |
interactive=False,
|
1567 |
)
|
1568 |
-
gr.Markdown(
|
1569 |
-
"""<p style="text-align: center;">② On the image, click 21 hand keypoints. This will serve as target hand poses. See the \"OpenPose keypoints convention\" for guidance.</p>"""
|
1570 |
-
)
|
1571 |
fix_kp_r_info = gr.Markdown(
|
1572 |
-
"""<p style="text-align: center;
|
1573 |
-
visible=False
|
1574 |
)
|
|
|
|
|
|
|
|
|
1575 |
fix_kp_right = gr.Image(
|
1576 |
type="numpy",
|
1577 |
label="Keypoint Selection (right hand)",
|
@@ -1590,7 +1863,7 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1590 |
value="Reset", interactive=False, visible=False
|
1591 |
)
|
1592 |
fix_kp_l_info = gr.Markdown(
|
1593 |
-
"""<p style="text-align: center;
|
1594 |
visible=False
|
1595 |
)
|
1596 |
fix_kp_left = gr.Image(
|
@@ -1621,13 +1894,15 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1621 |
width=LENGTH // 2,
|
1622 |
interactive=False,
|
1623 |
)
|
|
|
|
|
1624 |
with gr.Column():
|
1625 |
gr.Markdown(
|
1626 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">3. Press "Ready" to start pre-processing</p>"""
|
1627 |
)
|
1628 |
fix_ready = gr.Button(value="Ready", interactive=False)
|
1629 |
gr.Markdown(
|
1630 |
-
"""<p style="text-align: center; font-weight: bold; ">Visualized (256, 256)
|
1631 |
)
|
1632 |
fix_vis_mask32 = gr.Image(
|
1633 |
type="numpy",
|
@@ -1646,9 +1921,11 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1646 |
width=opts.image_size,
|
1647 |
interactive=False,
|
1648 |
)
|
1649 |
-
gr.Markdown(
|
1650 |
-
|
1651 |
-
)
|
|
|
|
|
1652 |
with gr.Column():
|
1653 |
gr.Markdown(
|
1654 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">4. Press "Run" to get the fixed hand image 🎯</p>"""
|
@@ -1657,6 +1934,16 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1657 |
gr.Markdown(
|
1658 |
"""<p style="text-align: center;">⚠️ >3min and ~24GB per generation</p>"""
|
1659 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1660 |
fix_result = gr.Gallery(
|
1661 |
type="numpy",
|
1662 |
label="Results",
|
@@ -1682,55 +1969,58 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1682 |
)
|
1683 |
fix_clear = gr.ClearButton()
|
1684 |
|
1685 |
-
gr.
|
1686 |
-
|
1687 |
-
|
1688 |
-
gr.Markdown(
|
1689 |
-
"⚠️ Currently, Number of generation > 1 could lead to out-of-memory"
|
1690 |
-
)
|
1691 |
-
with gr.Row():
|
1692 |
-
fix_n_generation = gr.Slider(
|
1693 |
-
label="Number of generations",
|
1694 |
-
value=1,
|
1695 |
-
minimum=1,
|
1696 |
-
maximum=FIX_MAX_N,
|
1697 |
-
step=1,
|
1698 |
-
randomize=False,
|
1699 |
-
interactive=True,
|
1700 |
-
)
|
1701 |
-
fix_seed = gr.Slider(
|
1702 |
-
label="Seed",
|
1703 |
-
value=42,
|
1704 |
-
minimum=0,
|
1705 |
-
maximum=10000,
|
1706 |
-
step=1,
|
1707 |
-
randomize=False,
|
1708 |
-
interactive=True,
|
1709 |
-
)
|
1710 |
-
fix_cfg = gr.Slider(
|
1711 |
-
label="Classifier free guidance scale",
|
1712 |
-
value=3.0,
|
1713 |
-
minimum=0.0,
|
1714 |
-
maximum=10.0,
|
1715 |
-
step=0.1,
|
1716 |
-
randomize=False,
|
1717 |
-
interactive=True,
|
1718 |
-
)
|
1719 |
-
fix_quality = gr.Slider(
|
1720 |
-
label="Quality",
|
1721 |
-
value=10,
|
1722 |
-
minimum=1,
|
1723 |
-
maximum=10,
|
1724 |
-
step=1,
|
1725 |
-
randomize=False,
|
1726 |
-
interactive=True,
|
1727 |
)
|
1728 |
-
|
1729 |
-
|
1730 |
-
|
1731 |
-
|
1732 |
-
|
1733 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1734 |
fix_img.change(lambda x: x, [fix_img], [fix_kp_right])
|
1735 |
fix_img.change(lambda x: x, [fix_img], [fix_kp_left])
|
1736 |
fix_inpaint_mask.change(
|
@@ -1775,7 +2065,7 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1775 |
],
|
1776 |
)
|
1777 |
fix_kp_right.select(
|
1778 |
-
get_kps, [fix_img, fix_kpts, gr.State("right")], [fix_kp_right, fix_kpts]
|
1779 |
)
|
1780 |
fix_undo_right.click(
|
1781 |
undo_kps, [fix_img, fix_kpts, gr.State("right")], [fix_kp_right, fix_kpts]
|
@@ -1797,7 +2087,7 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1797 |
)
|
1798 |
fix_ready.click(
|
1799 |
ready_sample,
|
1800 |
-
[
|
1801 |
[
|
1802 |
fix_ref_cond,
|
1803 |
fix_target_cond,
|
@@ -1816,23 +2106,28 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1816 |
fix_latent,
|
1817 |
fix_inpaint_latent,
|
1818 |
fix_kpts_np,
|
|
|
|
|
1819 |
fix_n_generation,
|
1820 |
fix_seed,
|
1821 |
fix_cfg,
|
1822 |
fix_quality,
|
1823 |
],
|
1824 |
-
[fix_result, fix_result_pose],
|
1825 |
)
|
1826 |
fix_clear.click(
|
1827 |
fix_clear_all,
|
1828 |
[],
|
1829 |
[
|
1830 |
fix_crop,
|
|
|
1831 |
fix_ref,
|
|
|
1832 |
fix_kp_right,
|
1833 |
fix_kp_left,
|
1834 |
fix_result,
|
1835 |
fix_result_pose,
|
|
|
1836 |
fix_inpaint_mask,
|
1837 |
fix_original,
|
1838 |
fix_img,
|
@@ -1850,6 +2145,20 @@ with gr.Blocks(css=custom_css, theme="soft") as demo:
|
|
1850 |
fix_quality,
|
1851 |
],
|
1852 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1853 |
|
1854 |
gr.Markdown("<h1>Citation</h1>")
|
1855 |
gr.Markdown(
|
|
|
20 |
from typing import Optional
|
21 |
import requests
|
22 |
from huggingface_hub import hf_hub_download
|
23 |
+
try:
|
24 |
+
import spaces
|
25 |
+
except:
|
26 |
+
pass
|
27 |
|
28 |
MAX_N = 6
|
29 |
FIX_MAX_N = 6
|
|
|
32 |
NEW_MODEL = True
|
33 |
MODEL_EPOCH = 6
|
34 |
REF_POSE_MASK = True
|
35 |
+
HF = False
|
36 |
+
pre_device = "cpu" if HF else "cuda"
|
37 |
+
spaces_60_fn = spaces.GPU(duration=60) if HF else (lambda f: f)
|
38 |
+
spaces_120_fn = spaces.GPU(duration=60) if HF else (lambda f: f)
|
39 |
+
spaces_300_fn = spaces.GPU(duration=60) if HF else (lambda f: f)
|
40 |
+
|
41 |
|
42 |
def set_seed(seed):
|
43 |
seed = int(seed)
|
|
|
121 |
# Convert BytesIO object to numpy array
|
122 |
buf.seek(0)
|
123 |
img_pil = Image.open(buf)
|
124 |
+
img_pil = img_pil.resize((W, H))
|
125 |
numpy_img = np.array(img_pil)
|
126 |
|
127 |
return numpy_img
|
|
|
241 |
print(f"encoder after eval() max: {max([p.max() for p in autoencoder.encoder.parameters()])}")
|
242 |
print(f"autoencoder encoder after eval() dtype: {next(autoencoder.encoder.parameters()).dtype}")
|
243 |
assert len(missing_keys) == 0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
244 |
|
245 |
+
sam_path = hf_hub_download(repo_id="Chaerin5/FoundHand-weights", filename="sam_vit_h_4b8939.pth", token=token)
|
246 |
+
sam_predictor = init_sam(ckpt_path=sam_path, device=pre_device)
|
247 |
|
248 |
print("Mediapipe hand detector and SAM ready...")
|
249 |
mp_hands = mp.solutions.hands
|
|
|
253 |
min_detection_confidence=0.1,
|
254 |
)
|
255 |
|
256 |
+
def prepare_anno(ref):
|
257 |
if ref is None:
|
258 |
return (
|
259 |
+
gr.Image.update(value=None),
|
260 |
+
gr.State.update(value=None),
|
|
|
|
|
|
|
261 |
)
|
|
|
|
|
262 |
img = ref["composite"][..., :3]
|
263 |
img = cv2.resize(img, opts.image_size, interpolation=cv2.INTER_AREA)
|
264 |
keypts = np.zeros((42, 2))
|
|
|
289 |
if keypts is None:
|
290 |
no_hands = cv2.resize(np.array(Image.open("no_hands.png"))[..., :3], (LENGTH, LENGTH))
|
291 |
return None, no_hands, None
|
292 |
+
missing_keys, extra_keys = autoencoder.load_state_dict(vae_state_dict, strict=False)
|
293 |
if isinstance(keypts, list):
|
294 |
if len(keypts[0]) == 0:
|
295 |
keypts[0] = np.zeros((21, 2))
|
|
|
298 |
else:
|
299 |
gr.Info("Number of right hand keypoints should be either 0 or 21.")
|
300 |
return None, None, None
|
|
|
301 |
if len(keypts[1]) == 0:
|
302 |
keypts[1] = np.zeros((21, 2))
|
303 |
elif len(keypts[1]) == 21:
|
|
|
305 |
else:
|
306 |
gr.Info("Number of left hand keypoints should be either 0 or 21.")
|
307 |
return None, None, None
|
|
|
308 |
keypts = np.concatenate(keypts, axis=0)
|
309 |
if REF_POSE_MASK:
|
310 |
sam_predictor.set_image(img)
|
|
|
343 |
Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
|
344 |
]
|
345 |
)
|
346 |
+
image = image_transform(img).to(device)
|
347 |
kpts_valid = check_keypoints_validity(keypts, target_size)
|
348 |
heatmaps = torch.tensor(
|
349 |
keypoint_heatmap(
|
|
|
351 |
)
|
352 |
* kpts_valid[:, None, None],
|
353 |
dtype=torch.float,
|
354 |
+
device=device
|
355 |
)[None, ...]
|
356 |
mask = torch.tensor(
|
357 |
cv2.resize(
|
|
|
360 |
interpolation=cv2.INTER_NEAREST,
|
361 |
),
|
362 |
dtype=torch.float,
|
363 |
+
device=device,
|
364 |
).unsqueeze(0)[None, ...]
|
365 |
return image[None, ...], heatmaps, mask
|
366 |
|
|
|
369 |
img,
|
370 |
keypts,
|
371 |
hand_mask,
|
372 |
+
device=pre_device,
|
373 |
target_size=opts.image_size,
|
374 |
latent_size=opts.latent_size,
|
375 |
)
|
|
|
390 |
|
391 |
return img, ref_pose, ref_cond
|
392 |
|
393 |
+
def get_target_anno(img, keypts):
|
394 |
+
if keypts is None:
|
395 |
+
no_hands = cv2.resize(np.array(Image.open("no_hands.png"))[..., :3], (LENGTH, LENGTH))
|
396 |
+
return None, no_hands, None, None
|
397 |
+
if isinstance(keypts, list):
|
398 |
+
if len(keypts[0]) == 0:
|
399 |
+
keypts[0] = np.zeros((21, 2))
|
400 |
+
elif len(keypts[0]) == 21:
|
401 |
+
keypts[0] = np.array(keypts[0], dtype=np.float32)
|
402 |
+
else:
|
403 |
+
gr.Info("Number of right hand keypoints should be either 0 or 21.")
|
404 |
+
return None, None, None
|
405 |
+
if len(keypts[1]) == 0:
|
406 |
+
keypts[1] = np.zeros((21, 2))
|
407 |
+
elif len(keypts[1]) == 21:
|
408 |
+
keypts[1] = np.array(keypts[1], dtype=np.float32)
|
409 |
+
else:
|
410 |
+
gr.Info("Number of left hand keypoints should be either 0 or 21.")
|
411 |
+
return None, None, None
|
412 |
+
keypts = np.concatenate(keypts, axis=0)
|
413 |
+
target_pose = visualize_hand(keypts, img)
|
414 |
+
kpts_valid = check_keypoints_validity(keypts, opts.image_size)
|
415 |
+
target_heatmaps = torch.tensor(
|
416 |
+
keypoint_heatmap(
|
417 |
+
scale_keypoint(keypts, opts.image_size, opts.latent_size),
|
418 |
+
opts.latent_size,
|
419 |
+
var=1.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
420 |
)
|
421 |
+
* kpts_valid[:, None, None],
|
422 |
+
dtype=torch.float,
|
423 |
+
device=pre_device,
|
424 |
+
)[None, ...]
|
425 |
+
target_cond = torch.cat(
|
426 |
+
[target_heatmaps, torch.zeros_like(target_heatmaps)[:, :1]], 1
|
427 |
+
)
|
428 |
|
429 |
+
return img, target_pose, target_cond, keypts
|
430 |
|
431 |
|
432 |
def get_mask_inpaint(ref):
|
433 |
+
# inpaint_mask = np.zeros_like(img_original[:, :, 0])
|
434 |
+
# cropped_mask = np.array(ref["layers"][0])[..., -1]
|
435 |
+
# inpaint_mask[crop_coord[0][1]:crop_coord[1][1], crop_coord[0][0]:crop_coord[1][0]] = cropped_mask
|
436 |
inpaint_mask = np.array(ref["layers"][0])[..., -1]
|
437 |
inpaint_mask = cv2.resize(
|
438 |
inpaint_mask, opts.image_size, interpolation=cv2.INTER_AREA
|
|
|
441 |
return inpaint_mask
|
442 |
|
443 |
|
444 |
+
def visualize_ref(brush): # crop,
|
445 |
+
if brush is None: # crop is None or
|
446 |
return None
|
447 |
inpainted = brush["layers"][0][..., -1]
|
448 |
+
img = brush["background"][..., :3]
|
449 |
+
# img = cv2.resize(img, inpainted.shape[::-1], interpolation=cv2.INTER_AREA)
|
450 |
mask = inpainted < 128
|
451 |
# img = img.astype(np.int32)
|
452 |
# img[mask, :] = img[mask, :] - 50
|
|
|
507 |
keypoints[1] = []
|
508 |
return img, keypoints
|
509 |
|
510 |
+
def stay_crop(img, crop_coord):
|
511 |
+
if img is not None:
|
512 |
+
crop_coord = [[0, 0], [img.shape[1], img.shape[0]]]
|
513 |
+
cropped = img.copy()
|
514 |
+
return crop_coord, cropped
|
515 |
+
else:
|
516 |
+
return None, None
|
517 |
+
|
518 |
+
def process_crop(img, crop_coord, evt:gr.SelectData):
|
519 |
+
if len(crop_coord) == 2:
|
520 |
+
crop_coord = [list(evt.index)]
|
521 |
+
cropped = img.copy()
|
522 |
+
elif len(crop_coord) == 1:
|
523 |
+
new_coord =list(evt.index)
|
524 |
+
if new_coord[0] <= crop_coord[0][0] or new_coord[1] <= crop_coord[0][1]:
|
525 |
+
gr.Warning("Second click should be more under and more right thand the first click. Try second click again.", duration=3)
|
526 |
+
cropped = img.copy()
|
527 |
+
else:
|
528 |
+
crop_coord.append(new_coord)
|
529 |
+
x1, y1 = crop_coord[0]
|
530 |
+
x2, y2 = crop_coord[1]
|
531 |
+
cropped = img.copy()[y1:y2, x1:x2]
|
532 |
+
else:
|
533 |
+
gr.Error("Something is wrong", duration=3)
|
534 |
+
return crop_coord, cropped
|
535 |
+
|
536 |
+
def disable_crop(crop_coord):
|
537 |
+
if len(crop_coord) == 2:
|
538 |
+
return gr.update(interactive=False)
|
539 |
+
else:
|
540 |
+
return gr.update(interactive=True)
|
541 |
+
|
542 |
+
@spaces_60_fn
|
543 |
def sample_diff(ref_cond, target_cond, target_keypts, num_gen, seed, cfg):
|
544 |
set_seed(seed)
|
545 |
z = torch.randn(
|
|
|
586 |
print(f"results[0].max(): {results[0].max()}")
|
587 |
return results, results_pose
|
588 |
|
589 |
+
@spaces_120_fn
|
590 |
+
def ready_sample(img_cropped, inpaint_mask, keypts):
|
591 |
+
# img = cv2.resize(img_ori[..., :3], opts.image_size, interpolation=cv2.INTER_AREA)
|
592 |
+
img = cv2.resize(img_cropped["background"][..., :3], opts.image_size, interpolation=cv2.INTER_AREA)
|
593 |
sam_predictor.set_image(img)
|
594 |
if len(keypts[0]) == 0:
|
595 |
keypts[0] = np.zeros((21, 2))
|
596 |
elif len(keypts[0]) == 21:
|
597 |
keypts[0] = np.array(keypts[0], dtype=np.float32)
|
598 |
+
# keypts[0][:, 0] = keypts[0][:, 0] + crop_coord[0][0]
|
599 |
+
# keypts[0][:, 1] = keypts[0][:, 1] + crop_coord[0][1]
|
600 |
else:
|
601 |
gr.Info("Number of right hand keypoints should be either 0 or 21.")
|
602 |
return None, None
|
|
|
605 |
keypts[1] = np.zeros((21, 2))
|
606 |
elif len(keypts[1]) == 21:
|
607 |
keypts[1] = np.array(keypts[1], dtype=np.float32)
|
608 |
+
# keypts[1][:, 0] = keypts[1][:, 0] + crop_coord[0][0]
|
609 |
+
# keypts[1][:, 1] = keypts[1][:, 1] + crop_coord[0][1]
|
610 |
else:
|
611 |
gr.Info("Number of left hand keypoints should be either 0 or 21.")
|
612 |
return None, None
|
613 |
|
614 |
keypts = np.concatenate(keypts, axis=0)
|
615 |
+
keypts = scale_keypoint(keypts, (img_cropped["background"].shape[1], img_cropped["background"].shape[0]), opts.image_size)
|
616 |
|
617 |
box_shift_ratio = 0.5
|
618 |
box_size_factor = 1.2
|
|
|
648 |
inpaint_mask, dsize=opts.latent_size, interpolation=cv2.INTER_NEAREST
|
649 |
),
|
650 |
dtype=torch.float,
|
651 |
+
device=pre_device,
|
652 |
).unsqueeze(0)[None, ...]
|
653 |
|
654 |
def make_ref_cond(
|
|
|
666 |
Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
|
667 |
]
|
668 |
)
|
669 |
+
image = image_transform(img).to(device)
|
670 |
kpts_valid = check_keypoints_validity(keypts, target_size)
|
671 |
heatmaps = torch.tensor(
|
672 |
keypoint_heatmap(
|
|
|
674 |
)
|
675 |
* kpts_valid[:, None, None],
|
676 |
dtype=torch.float,
|
677 |
+
device=device,
|
678 |
)[None, ...]
|
679 |
mask = torch.tensor(
|
680 |
cv2.resize(
|
|
|
683 |
interpolation=cv2.INTER_NEAREST,
|
684 |
),
|
685 |
dtype=torch.float,
|
686 |
+
device=device,
|
687 |
).unsqueeze(0)[None, ...]
|
688 |
return image[None, ...], heatmaps, mask
|
689 |
|
|
|
691 |
img,
|
692 |
keypts,
|
693 |
hand_mask * (1 - inpaint_mask),
|
694 |
+
device=pre_device,
|
695 |
target_size=opts.image_size,
|
696 |
latent_size=opts.latent_size,
|
697 |
)
|
|
|
731 |
out = (gr.update(visible=True), gr.update(visible=False))
|
732 |
return out
|
733 |
|
734 |
+
@spaces_300_fn
|
735 |
def sample_inpaint(
|
736 |
ref_cond,
|
737 |
target_cond,
|
738 |
latent,
|
739 |
inpaint_latent_mask,
|
740 |
keypts,
|
741 |
+
img_original,
|
742 |
+
crop_coord,
|
743 |
num_gen,
|
744 |
seed,
|
745 |
cfg,
|
|
|
785 |
# visualize
|
786 |
results = []
|
787 |
results_pose = []
|
788 |
+
results_original = []
|
789 |
for i in range(FIX_MAX_N):
|
790 |
if i < num_gen:
|
791 |
+
res =sampled_images[i]
|
792 |
+
results.append(res)
|
793 |
+
results_pose.append(visualize_hand(keypts, res))
|
794 |
+
res = cv2.resize(res, (crop_coord[1][0]-crop_coord[0][0], crop_coord[1][1]-crop_coord[0][1]))
|
795 |
+
res_original = img_original.copy()
|
796 |
+
res_original[crop_coord[0][1]:crop_coord[1][1], crop_coord[0][0]:crop_coord[1][0], :] = res
|
797 |
+
results_original.append(res_original)
|
798 |
else:
|
799 |
results.append(placeholder)
|
800 |
results_pose.append(placeholder)
|
801 |
+
results_original.append(placeholder)
|
802 |
+
return results, results_pose, results_original
|
803 |
|
804 |
|
805 |
def flip_hand(
|
806 |
+
img, img_raw, pose_img, pose_manual_img,
|
807 |
+
manual_kp_right, manual_kp_left,
|
808 |
+
cond, auto_cond, manual_cond,
|
809 |
+
keypts=None, auto_keypts=None, manual_keypts=None
|
810 |
):
|
811 |
if cond is None: # clear clicked
|
812 |
+
return
|
813 |
img["composite"] = img["composite"][:, ::-1, :]
|
814 |
img["background"] = img["background"][:, ::-1, :]
|
815 |
img["layers"] = [layer[:, ::-1, :] for layer in img["layers"]]
|
816 |
+
# for comp in [pose_img, pose_manual_img, manual_kp_right, manual_kp_left, cond, auto_cond, manual_cond]:
|
817 |
+
# if comp is not None:
|
818 |
+
# if isinstance(comp, torch.Tensor):
|
819 |
+
# comp = comp.flip(-1)
|
820 |
+
# else:
|
821 |
+
# comp = comp[:, ::-1, :]
|
822 |
+
if img_raw is not None:
|
823 |
+
img_raw = img_raw[:, ::-1, :]
|
824 |
pose_img = pose_img[:, ::-1, :]
|
825 |
+
if pose_manual_img is not None:
|
826 |
+
pose_manual_img = pose_manual_img[:, ::-1, :]
|
827 |
+
if manual_kp_right is not None:
|
828 |
+
manual_kp_right = manual_kp_right[:, ::-1, :]
|
829 |
+
if manual_kp_left is not None:
|
830 |
+
manual_kp_left = manual_kp_left[:, ::-1, :]
|
831 |
cond = cond.flip(-1)
|
832 |
+
if auto_cond is not None:
|
833 |
+
auto_cond = auto_cond.flip(-1)
|
834 |
+
if manual_cond is not None:
|
835 |
+
manual_cond = manual_cond.flip(-1)
|
836 |
+
# for comp in [keypts, auto_keypts, manual_keypts]:
|
837 |
+
# if comp is not None:
|
838 |
+
# if comp[:21, :].sum() != 0:
|
839 |
+
# comp[:21, 0] = opts.image_size[1] - comp[:21, 0]
|
840 |
+
# if comp[21:, :].sum() != 0:
|
841 |
+
# comp[21:, 0] = opts.image_size[1] - comp[21:, 0]
|
842 |
+
if keypts is not None:
|
843 |
if keypts[:21, :].sum() != 0:
|
844 |
keypts[:21, 0] = opts.image_size[1] - keypts[:21, 0]
|
|
|
845 |
if keypts[21:, :].sum() != 0:
|
846 |
keypts[21:, 0] = opts.image_size[1] - keypts[21:, 0]
|
847 |
+
if auto_keypts is not None:
|
848 |
+
if auto_keypts[:21, :].sum() != 0:
|
849 |
+
auto_keypts[:21, 0] = opts.image_size[1] - auto_keypts[:21, 0]
|
850 |
+
if auto_keypts[21:, :].sum() != 0:
|
851 |
+
auto_keypts[21:, 0] = opts.image_size[1] - auto_keypts[21:, 0]
|
852 |
+
if manual_keypts is not None:
|
853 |
+
if manual_keypts[:21, :].sum() != 0:
|
854 |
+
manual_keypts[:21, 0] = opts.image_size[1] - manual_keypts[:21, 0]
|
855 |
+
if manual_keypts[21:, :].sum() != 0:
|
856 |
+
manual_keypts[21:, 0] = opts.image_size[1] - manual_keypts[21:, 0]
|
857 |
+
return img, img_raw, pose_img, pose_manual_img, manual_kp_right, manual_kp_left, cond, auto_cond, manual_cond, keypts, auto_keypts, manual_keypts
|
858 |
|
859 |
|
860 |
def resize_to_full(img):
|
|
|
867 |
def clear_all():
|
868 |
return (
|
869 |
None,
|
870 |
+
[],
|
871 |
None,
|
872 |
None,
|
873 |
None,
|
874 |
None,
|
|
|
875 |
None,
|
876 |
None,
|
877 |
False,
|
878 |
None,
|
879 |
None,
|
880 |
+
[],
|
881 |
+
None,
|
882 |
+
None,
|
883 |
+
None,
|
884 |
None,
|
885 |
None,
|
886 |
None,
|
887 |
+
False,
|
888 |
None,
|
889 |
None,
|
890 |
1,
|
891 |
42,
|
892 |
3.0,
|
893 |
gr.update(interactive=False),
|
|
|
894 |
)
|
895 |
|
896 |
|
|
|
899 |
None,
|
900 |
None,
|
901 |
None,
|
902 |
+
[],
|
903 |
+
None,
|
904 |
+
None,
|
905 |
None,
|
906 |
None,
|
907 |
None,
|
|
|
927 |
def enable_component(image1, image2):
|
928 |
if image1 is None or image2 is None:
|
929 |
return gr.update(interactive=False)
|
930 |
+
if isinstance(image1, dict) and "background" in image1 and "layers" in image1 and "composite" in image1:
|
931 |
if (
|
932 |
image1["background"].sum() == 0
|
933 |
and (sum([im.sum() for im in image1["layers"]]) == 0)
|
934 |
and image1["composite"].sum() == 0
|
935 |
):
|
936 |
return gr.update(interactive=False)
|
937 |
+
if isinstance(image1, dict) and "background" in image2 and "layers" in image2 and "composite" in image2:
|
938 |
if (
|
939 |
image2["background"].sum() == 0
|
940 |
and (sum([im.sum() for im in image2["layers"]]) == 0)
|
|
|
991 |
|
992 |
def set_unvisible():
|
993 |
return (
|
994 |
+
gr.update(visible=False),
|
995 |
+
gr.update(visible=False),
|
996 |
+
gr.update(visible=False),
|
997 |
+
gr.update(visible=False),
|
998 |
+
gr.update(visible=False),
|
999 |
+
gr.update(visible=False),
|
1000 |
+
gr.update(visible=False),
|
1001 |
+
gr.update(visible=False),
|
1002 |
+
gr.update(visible=False),
|
1003 |
+
gr.update(visible=False),
|
1004 |
+
gr.update(visible=False),
|
1005 |
+
gr.update(visible=False),
|
1006 |
gr.update(visible=False),
|
1007 |
gr.update(visible=False),
|
1008 |
gr.update(visible=False),
|
|
|
1017 |
gr.update(visible=False)
|
1018 |
)
|
1019 |
|
1020 |
+
def fix_set_unvisible():
|
1021 |
+
return (
|
1022 |
+
gr.update(visible=False),
|
1023 |
+
gr.update(visible=False),
|
1024 |
+
gr.update(visible=False),
|
1025 |
+
gr.update(visible=False),
|
1026 |
+
gr.update(visible=False),
|
1027 |
+
gr.update(visible=False),
|
1028 |
+
gr.update(visible=False),
|
1029 |
+
gr.update(visible=False)
|
1030 |
+
)
|
1031 |
+
|
1032 |
def set_no_hands(decider, component):
|
1033 |
if decider is None:
|
1034 |
no_hands = cv2.resize(np.array(Image.open("no_hands.png"))[..., :3], (LENGTH, LENGTH))
|
|
|
1050 |
update_component = gr.update(visible=True)
|
1051 |
return update_component
|
1052 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1053 |
LENGTH = 480
|
1054 |
|
1055 |
example_ref_imgs = [
|
|
|
1145 |
# ["bad_hands/4.jpg"], # "bad_hands/4_mask.jpg"],
|
1146 |
["bad_hands/5.jpg"], # "bad_hands/5_mask.jpg"],
|
1147 |
["bad_hands/6.jpg"], # "bad_hands/6_mask.jpg"],
|
1148 |
+
# ["bad_hands/7.jpg"], # "bad_hands/7_mask.jpg"],
|
1149 |
# ["bad_hands/8.jpg"], # "bad_hands/8_mask.jpg"],
|
1150 |
# ["bad_hands/9.jpg"], # "bad_hands/9_mask.jpg"],
|
1151 |
# ["bad_hands/10.jpg"], # "bad_hands/10_mask.jpg"],
|
|
|
1199 |
with gr.Blocks(css=custom_css, theme="soft") as demo:
|
1200 |
gr.Markdown(_HEADER_)
|
1201 |
with gr.Tab("Edit Hand Poses"):
|
1202 |
+
dump = gr.State(value=None)
|
1203 |
+
|
1204 |
+
# ref states
|
1205 |
ref_img = gr.State(value=None)
|
1206 |
ref_im_raw = gr.State(value=None)
|
1207 |
ref_kp_raw = gr.State(value=0)
|
1208 |
ref_kp_got = gr.State(value=None)
|
|
|
|
|
1209 |
ref_manual_cond = gr.State(value=None)
|
1210 |
ref_auto_cond = gr.State(value=None)
|
1211 |
+
ref_cond = gr.State(value=None)
|
1212 |
+
|
1213 |
+
# target states
|
1214 |
target_img = gr.State(value=None)
|
1215 |
+
target_im_raw = gr.State(value=None)
|
1216 |
+
target_kp_raw = gr.State(value=0)
|
1217 |
+
target_kp_got = gr.State(value=None)
|
1218 |
+
target_manual_keypts = gr.State(value=None)
|
1219 |
+
target_auto_keypts = gr.State(value=None)
|
1220 |
target_keypts = gr.State(value=None)
|
1221 |
+
target_manual_cond = gr.State(value=None)
|
1222 |
+
target_auto_cond = gr.State(value=None)
|
1223 |
+
target_cond = gr.State(value=None)
|
1224 |
+
|
1225 |
+
# main tab
|
1226 |
with gr.Row():
|
1227 |
+
# ref column
|
1228 |
with gr.Column():
|
1229 |
gr.Markdown(
|
1230 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">1. Upload a hand image to edit 📥</p>"""
|
|
|
1344 |
ref_flip = gr.Checkbox(
|
1345 |
value=False, label="Flip Handedness (Reference)", interactive=False
|
1346 |
)
|
1347 |
+
|
1348 |
+
# target column
|
1349 |
with gr.Column():
|
1350 |
gr.Markdown(
|
1351 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">2. Upload a hand image for target hand pose 📥</p>"""
|
|
|
1370 |
target_finish_crop = gr.Button(
|
1371 |
value="Finish Cropping", interactive=False
|
1372 |
)
|
1373 |
+
with gr.Tab("Automatic hand keypoints"):
|
1374 |
+
target_pose = gr.Image(
|
1375 |
+
type="numpy",
|
1376 |
+
label="Target Pose",
|
1377 |
+
show_label=True,
|
1378 |
+
height=LENGTH,
|
1379 |
+
width=LENGTH,
|
1380 |
+
interactive=False,
|
1381 |
+
)
|
1382 |
+
target_use_auto = gr.Button(value="Click here to use automatic, not manual", interactive=False, visible=True)
|
1383 |
+
with gr.Tab("Manual hand keypoints"):
|
1384 |
+
target_manual_checkbox_info = gr.Markdown(
|
1385 |
+
"""<p style="text-align: center;"><b>Step 1.</b> Tell us if this is right, left, or both hands.</p>""",
|
1386 |
+
visible=True,
|
1387 |
+
)
|
1388 |
+
target_manual_checkbox = gr.CheckboxGroup(
|
1389 |
+
["Right hand", "Left hand"],
|
1390 |
+
show_label=False,
|
1391 |
+
visible=True,
|
1392 |
+
interactive=True,
|
1393 |
+
)
|
1394 |
+
target_manual_kp_r_info = gr.Markdown(
|
1395 |
+
"""<p style="text-align: center;"><b>Step 2.</b> Click on image to provide hand keypoints for <b>right</b> hand. See \"OpenPose Keypoint Convention\" for guidance.</p>""",
|
1396 |
+
visible=False,
|
1397 |
+
)
|
1398 |
+
target_manual_kp_right = gr.Image(
|
1399 |
+
type="numpy",
|
1400 |
+
label="Keypoint Selection (right hand)",
|
1401 |
+
show_label=True,
|
1402 |
+
height=LENGTH,
|
1403 |
+
width=LENGTH,
|
1404 |
+
interactive=False,
|
1405 |
+
visible=False,
|
1406 |
+
sources=[],
|
1407 |
+
)
|
1408 |
+
with gr.Row():
|
1409 |
+
target_manual_undo_right = gr.Button(
|
1410 |
+
value="Undo", interactive=True, visible=False
|
1411 |
+
)
|
1412 |
+
target_manual_reset_right = gr.Button(
|
1413 |
+
value="Reset", interactive=True, visible=False
|
1414 |
+
)
|
1415 |
+
target_manual_kp_l_info = gr.Markdown(
|
1416 |
+
"""<p style="text-align: center;"><b>Step 2.</b> Click on image to provide hand keypoints for <b>left</b> hand. See \"OpenPose keypoint convention\" for guidance.</p>""",
|
1417 |
+
visible=False
|
1418 |
+
)
|
1419 |
+
target_manual_kp_left = gr.Image(
|
1420 |
+
type="numpy",
|
1421 |
+
label="Keypoint Selection (left hand)",
|
1422 |
+
show_label=True,
|
1423 |
+
height=LENGTH,
|
1424 |
+
width=LENGTH,
|
1425 |
+
interactive=False,
|
1426 |
+
visible=False,
|
1427 |
+
sources=[],
|
1428 |
+
)
|
1429 |
+
with gr.Row():
|
1430 |
+
target_manual_undo_left = gr.Button(
|
1431 |
+
value="Undo", interactive=True, visible=False
|
1432 |
+
)
|
1433 |
+
target_manual_reset_left = gr.Button(
|
1434 |
+
value="Reset", interactive=True, visible=False
|
1435 |
+
)
|
1436 |
+
target_manual_done_info = gr.Markdown(
|
1437 |
+
"""<p style="text-align: center;"><b>Step 3.</b> Hit \"Done\" button to confirm.</p>""",
|
1438 |
+
visible=False,
|
1439 |
+
)
|
1440 |
+
target_manual_done = gr.Button(value="Done", interactive=True, visible=False)
|
1441 |
+
target_manual_pose = gr.Image(
|
1442 |
+
type="numpy",
|
1443 |
+
label="Target Pose",
|
1444 |
+
show_label=True,
|
1445 |
+
height=LENGTH,
|
1446 |
+
width=LENGTH,
|
1447 |
+
interactive=False,
|
1448 |
+
visible=False
|
1449 |
+
)
|
1450 |
+
target_use_manual = gr.Button(value="Click here to use manual, not automatic", interactive=True, visible=False)
|
1451 |
+
target_manual_instruct = gr.Markdown(
|
1452 |
+
value="""<p style="text-align: left; font-weight: bold; ">OpenPose Keypoints Convention</p>""",
|
1453 |
+
visible=True
|
1454 |
+
)
|
1455 |
+
target_manual_openpose = gr.Image(
|
1456 |
+
value="openpose.png",
|
1457 |
+
type="numpy",
|
1458 |
+
show_label=False,
|
1459 |
+
height=LENGTH // 2,
|
1460 |
+
width=LENGTH // 2,
|
1461 |
+
interactive=False,
|
1462 |
+
visible=True
|
1463 |
+
)
|
1464 |
gr.Markdown(
|
1465 |
"""<p style="text-align: center;">③ Optionally flip the hand</p>"""
|
1466 |
)
|
1467 |
target_flip = gr.Checkbox(
|
1468 |
value=False, label="Flip Handedness (Target)", interactive=False
|
1469 |
)
|
1470 |
+
|
1471 |
+
# result column
|
1472 |
with gr.Column():
|
1473 |
gr.Markdown(
|
1474 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">3. Press "Run" to get the edited results 🎯</p>"""
|
|
|
1532 |
interactive=True,
|
1533 |
)
|
1534 |
|
1535 |
+
# reference listeners
|
1536 |
ref.change(enable_component, [ref, ref], ref_finish_crop)
|
1537 |
+
ref_finish_crop.click(prepare_anno, [ref], [ref_im_raw, ref_kp_raw])
|
1538 |
ref_kp_raw.change(lambda x: x, ref_im_raw, ref_manual_kp_right)
|
1539 |
ref_kp_raw.change(lambda x: x, ref_im_raw, ref_manual_kp_left)
|
1540 |
+
ref_kp_raw.change(get_ref_anno, [ref_im_raw, ref_kp_raw], [ref_img, ref_pose, ref_auto_cond])
|
1541 |
+
ref_pose.change(enable_component, [ref_kp_raw, ref_pose], ref_use_auto)
|
1542 |
+
ref_pose.change(enable_component, [ref_img, ref_pose], ref_flip)
|
1543 |
+
ref_auto_cond.change(lambda x: x, ref_auto_cond, ref_cond)
|
1544 |
+
ref_use_auto.click(lambda x: x, ref_auto_cond, ref_cond)
|
1545 |
+
ref_use_auto.click(lambda x: gr.Info("Automatic hand keypoints will be used for 'Reference'", duration=3))
|
1546 |
+
|
1547 |
ref_manual_checkbox.select(
|
1548 |
set_visible,
|
1549 |
[ref_manual_checkbox, ref_kp_got, ref_im_raw, ref_manual_kp_right, ref_manual_kp_left, ref_manual_done],
|
|
|
1581 |
ref_manual_reset_left.click(
|
1582 |
reset_kps, [ref_im_raw, ref_kp_got, gr.State("left")], [ref_manual_kp_left, ref_kp_got]
|
1583 |
)
|
1584 |
+
ref_manual_done.click(visible_component, [gr.State(0), ref_manual_pose], ref_manual_pose)
|
1585 |
+
ref_manual_done.click(visible_component, [gr.State(0), ref_use_manual], ref_use_manual)
|
1586 |
ref_manual_done.click(get_ref_anno, [ref_im_raw, ref_kp_got], [ref_img, ref_manual_pose, ref_manual_cond])
|
|
|
|
|
|
|
|
|
|
|
1587 |
ref_manual_pose.change(enable_component, [ref_manual_pose, ref_manual_pose], ref_manual_done)
|
|
|
|
|
|
|
|
|
|
|
|
|
1588 |
ref_manual_pose.change(enable_component, [ref_img, ref_manual_pose], ref_flip)
|
1589 |
+
ref_manual_cond.change(lambda x: x, ref_manual_cond, ref_cond)
|
1590 |
+
ref_use_manual.click(lambda x: x, ref_manual_cond, ref_cond)
|
1591 |
+
ref_use_manual.click(lambda x: gr.Info("Manual hand keypoints will be used for 'Reference'", duration=3))
|
1592 |
+
|
1593 |
ref_flip.select(
|
1594 |
+
flip_hand,
|
1595 |
+
[ref, ref_im_raw, ref_pose, ref_manual_pose, ref_manual_kp_right, ref_manual_kp_left, ref_cond, ref_auto_cond, ref_manual_cond],
|
1596 |
+
[ref, ref_im_raw, ref_pose, ref_manual_pose, ref_manual_kp_right, ref_manual_kp_left, ref_cond, ref_auto_cond, ref_manual_cond]
|
1597 |
)
|
1598 |
+
|
1599 |
+
# target listeners
|
1600 |
target.change(enable_component, [target, target], target_finish_crop)
|
1601 |
+
target_finish_crop.click(prepare_anno, [target], [target_im_raw, target_kp_raw])
|
1602 |
+
target_kp_raw.change(lambda x:x, target_im_raw, target_manual_kp_right)
|
1603 |
+
target_kp_raw.change(lambda x:x, target_im_raw, target_manual_kp_left)
|
1604 |
+
target_kp_raw.change(get_target_anno, [target_im_raw, target_kp_raw], [target_img, target_pose, target_auto_cond, target_auto_keypts])
|
1605 |
+
target_pose.change(enable_component, [target_kp_raw, target_pose], target_use_auto)
|
1606 |
target_pose.change(enable_component, [target_img, target_pose], target_flip)
|
1607 |
+
target_auto_cond.change(lambda x: x, target_auto_cond, target_cond)
|
1608 |
+
target_auto_keypts.change(lambda x: x, target_auto_keypts, target_keypts)
|
1609 |
+
target_use_auto.click(lambda x: x, target_auto_cond, target_cond)
|
1610 |
+
target_use_auto.click(lambda x: x, target_auto_keypts, target_keypts)
|
1611 |
+
target_use_auto.click(lambda x: gr.Info("Automatic hand keypoints will be used for 'Target'", duration=3))
|
1612 |
+
|
1613 |
+
target_manual_checkbox.select(
|
1614 |
+
set_visible,
|
1615 |
+
[target_manual_checkbox, target_kp_got, target_im_raw, target_manual_kp_right, target_manual_kp_left, target_manual_done],
|
1616 |
+
[
|
1617 |
+
target_kp_got,
|
1618 |
+
target_manual_kp_right,
|
1619 |
+
target_manual_kp_left,
|
1620 |
+
target_manual_kp_right,
|
1621 |
+
target_manual_undo_right,
|
1622 |
+
target_manual_reset_right,
|
1623 |
+
target_manual_kp_left,
|
1624 |
+
target_manual_undo_left,
|
1625 |
+
target_manual_reset_left,
|
1626 |
+
target_manual_kp_r_info,
|
1627 |
+
target_manual_kp_l_info,
|
1628 |
+
target_manual_done,
|
1629 |
+
target_manual_done_info
|
1630 |
+
]
|
1631 |
+
)
|
1632 |
+
target_manual_kp_right.select(
|
1633 |
+
get_kps, [target_im_raw, target_kp_got, gr.State("right")], [target_manual_kp_right, target_kp_got]
|
1634 |
+
)
|
1635 |
+
target_manual_undo_right.click(
|
1636 |
+
undo_kps, [target_im_raw, target_kp_got, gr.State("right")], [target_manual_kp_right, target_kp_got]
|
1637 |
+
)
|
1638 |
+
target_manual_reset_right.click(
|
1639 |
+
reset_kps, [target_im_raw, target_kp_got, gr.State("right")], [target_manual_kp_right, target_kp_got]
|
1640 |
+
)
|
1641 |
+
target_manual_kp_left.select(
|
1642 |
+
get_kps, [target_im_raw, target_kp_got, gr.State("left")], [target_manual_kp_left, target_kp_got]
|
1643 |
+
)
|
1644 |
+
target_manual_undo_left.click(
|
1645 |
+
undo_kps, [target_im_raw, target_kp_got, gr.State("left")], [target_manual_kp_left, target_kp_got]
|
1646 |
+
)
|
1647 |
+
target_manual_reset_left.click(
|
1648 |
+
reset_kps, [target_im_raw, target_kp_got, gr.State("left")], [target_manual_kp_left, target_kp_got]
|
1649 |
+
)
|
1650 |
+
target_manual_done.click(visible_component, [gr.State(0), target_manual_pose], target_manual_pose)
|
1651 |
+
target_manual_done.click(visible_component, [gr.State(0), target_use_manual], target_use_manual)
|
1652 |
+
target_manual_done.click(get_target_anno, [target_im_raw, target_kp_got], [target_img, target_manual_pose, target_manual_cond, target_manual_keypts])
|
1653 |
+
target_manual_pose.change(enable_component, [target_manual_pose, target_manual_pose], target_manual_done)
|
1654 |
+
target_manual_pose.change(enable_component, [target_img, target_manual_pose], target_flip)
|
1655 |
+
target_manual_cond.change(lambda x: x, target_manual_cond, target_cond)
|
1656 |
+
target_manual_keypts.change(lambda x: x, target_manual_keypts, target_keypts)
|
1657 |
+
target_use_manual.click(lambda x: x, target_manual_cond, target_cond)
|
1658 |
+
target_use_manual.click(lambda x: x, target_manual_keypts, target_keypts)
|
1659 |
+
target_use_manual.click(lambda x: gr.Info("Manual hand keypoints will be used for 'Reference'", duration=3))
|
1660 |
+
|
1661 |
target_flip.select(
|
1662 |
flip_hand,
|
1663 |
+
[target, target_im_raw, target_pose, target_manual_pose, target_manual_kp_right, target_manual_kp_left, target_cond, target_auto_cond, target_manual_cond, target_keypts, target_auto_keypts, target_manual_keypts],
|
1664 |
+
[target, target_im_raw, target_pose, target_manual_pose, target_manual_kp_right, target_manual_kp_left, target_cond, target_auto_cond, target_manual_cond, target_keypts, target_auto_keypts, target_manual_keypts],
|
1665 |
)
|
1666 |
+
|
1667 |
+
# run listerners
|
1668 |
+
ref_cond.change(enable_component, [ref_cond, target_cond], run)
|
1669 |
+
target_cond.change(enable_component, [ref_cond, target_cond], run)
|
1670 |
+
# ref_manual_pose.change(enable_component, [ref_manual_pose, target_manual_pose], run)
|
1671 |
+
# target_manual_pose.change(enable_component, [ref_manual_pose, target_manual_pose], run)
|
1672 |
run.click(
|
1673 |
sample_diff,
|
1674 |
[ref_cond, target_cond, target_keypts, n_generation, seed, cfg],
|
|
|
1679 |
[],
|
1680 |
[
|
1681 |
ref,
|
1682 |
+
ref_manual_checkbox,
|
1683 |
ref_manual_kp_right,
|
1684 |
ref_manual_kp_left,
|
1685 |
+
ref_img,
|
1686 |
ref_pose,
|
1687 |
ref_manual_pose,
|
1688 |
+
ref_cond,
|
1689 |
ref_flip,
|
1690 |
target,
|
1691 |
+
target_keypts,
|
1692 |
+
target_manual_checkbox,
|
1693 |
+
target_manual_kp_right,
|
1694 |
+
target_manual_kp_left,
|
1695 |
+
target_img,
|
1696 |
target_pose,
|
1697 |
+
target_manual_pose,
|
1698 |
+
target_cond,
|
1699 |
target_flip,
|
1700 |
results,
|
1701 |
results_pose,
|
|
|
|
|
|
|
|
|
|
|
1702 |
n_generation,
|
1703 |
seed,
|
1704 |
cfg,
|
1705 |
ref_kp_raw,
|
|
|
1706 |
],
|
1707 |
)
|
1708 |
clear.click(
|
1709 |
set_unvisible,
|
1710 |
[],
|
1711 |
[
|
|
|
1712 |
ref_manual_kp_l_info,
|
1713 |
+
ref_manual_kp_r_info,
|
1714 |
+
ref_manual_kp_left,
|
1715 |
+
ref_manual_kp_right,
|
1716 |
ref_manual_undo_left,
|
1717 |
ref_manual_undo_right,
|
1718 |
ref_manual_reset_left,
|
|
|
1721 |
ref_manual_done_info,
|
1722 |
ref_manual_pose,
|
1723 |
ref_use_manual,
|
1724 |
+
target_manual_kp_l_info,
|
1725 |
+
target_manual_kp_r_info,
|
1726 |
+
target_manual_kp_left,
|
1727 |
+
target_manual_kp_right,
|
1728 |
+
target_manual_undo_left,
|
1729 |
+
target_manual_undo_right,
|
1730 |
+
target_manual_reset_left,
|
1731 |
+
target_manual_reset_right,
|
1732 |
+
target_manual_done,
|
1733 |
+
target_manual_done_info,
|
1734 |
+
target_manual_pose,
|
1735 |
+
target_use_manual,
|
1736 |
]
|
1737 |
)
|
1738 |
|
1739 |
with gr.Tab("Fix Hands"):
|
1740 |
fix_inpaint_mask = gr.State(value=None)
|
1741 |
fix_original = gr.State(value=None)
|
1742 |
+
fix_crop_coord = gr.State(value=None)
|
1743 |
fix_img = gr.State(value=None)
|
1744 |
fix_kpts = gr.State(value=None)
|
1745 |
fix_kpts_np = gr.State(value=None)
|
|
|
1748 |
fix_latent = gr.State(value=None)
|
1749 |
fix_inpaint_latent = gr.State(value=None)
|
1750 |
with gr.Row():
|
1751 |
+
# crop & brush
|
1752 |
with gr.Column():
|
1753 |
gr.Markdown(
|
1754 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">1. Upload a malformed hand image to fix 📥</p>"""
|
1755 |
)
|
1756 |
gr.Markdown(
|
1757 |
+
"""<p style="text-align: center;">① Optionally crop the image by clicking <b>top left</b> and <b>bottom right</b> of your desired bounding box around the hand. </p>"""
|
1758 |
)
|
1759 |
+
# fix_crop = gr.ImageEditor(
|
1760 |
+
# type="numpy",
|
1761 |
+
# sources=["upload", "webcam", "clipboard"],
|
1762 |
+
# label="Image crop",
|
1763 |
+
# show_label=True,
|
1764 |
+
# height=LENGTH,
|
1765 |
+
# width=LENGTH,
|
1766 |
+
# layers=False,
|
1767 |
+
# # crop_size="1:1",
|
1768 |
+
# transforms=(),
|
1769 |
+
# brush=False,
|
1770 |
+
# image_mode="RGBA",
|
1771 |
+
# container=False,
|
1772 |
+
# )
|
1773 |
+
fix_crop = gr.Image(
|
1774 |
type="numpy",
|
1775 |
sources=["upload", "webcam", "clipboard"],
|
1776 |
+
label="Input Image",
|
1777 |
show_label=True,
|
1778 |
height=LENGTH,
|
1779 |
width=LENGTH,
|
1780 |
+
interactive=True,
|
1781 |
+
visible=True,
|
1782 |
+
)
|
1783 |
+
gr.Markdown(
|
1784 |
+
"""<p style="text-align: center;">💡 If you crop, the model can focus on more details of the cropped area. Square crops might work better than rectangle crops.</p>"""
|
1785 |
)
|
1786 |
+
# fix_tmp = gr.Image(
|
1787 |
+
# type="numpy",
|
1788 |
+
# label="tmp",
|
1789 |
+
# show_label=True,
|
1790 |
+
# height=LENGTH,
|
1791 |
+
# width=LENGTH,
|
1792 |
+
# interactive=True,
|
1793 |
+
# visible=True,
|
1794 |
+
# sources=[],
|
1795 |
+
# )
|
1796 |
fix_example = gr.Examples(
|
1797 |
fix_example_imgs,
|
1798 |
inputs=[fix_crop],
|
1799 |
examples_per_page=20,
|
1800 |
)
|
1801 |
gr.Markdown(
|
1802 |
+
"""<p style="text-align: center;">② Brush area (e.g., wrong finger) that needs to be fixed. Don't brush the entire hand!</p>"""
|
1803 |
)
|
1804 |
fix_ref = gr.ImageEditor(
|
1805 |
type="numpy",
|
1806 |
+
label="Image Brushing",
|
1807 |
sources=(),
|
1808 |
show_label=True,
|
1809 |
height=LENGTH,
|
|
|
1817 |
container=False,
|
1818 |
interactive=False,
|
1819 |
)
|
1820 |
+
gr.Markdown(
|
1821 |
+
"""<p style="text-align: center;">③ Hit the \"Finish Cropping & Brushing\" button</p>"""
|
1822 |
+
)
|
1823 |
fix_finish_crop = gr.Button(
|
1824 |
value="Finish Croping & Brushing", interactive=False
|
1825 |
)
|
1826 |
+
|
1827 |
+
# keypoint selection
|
1828 |
with gr.Column():
|
1829 |
gr.Markdown(
|
1830 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">2. Click on hand to get target hand pose</p>"""
|
|
|
1837 |
show_label=False,
|
1838 |
interactive=False,
|
1839 |
)
|
|
|
|
|
|
|
1840 |
fix_kp_r_info = gr.Markdown(
|
1841 |
+
"""<p style="text-align: center;">② Click 21 keypoints on the image to provide the target hand pose of <b>right hand</b>. See the \"OpenPose keypoints convention\" for guidance.</p>""",
|
1842 |
+
visible=False
|
1843 |
)
|
1844 |
+
# fix_kp_r_info = gr.Markdown(
|
1845 |
+
# """<p style="text-align: center; font-size: 20px; font-weight: bold; ">Select right only</p>""",
|
1846 |
+
# visible=False,
|
1847 |
+
# )
|
1848 |
fix_kp_right = gr.Image(
|
1849 |
type="numpy",
|
1850 |
label="Keypoint Selection (right hand)",
|
|
|
1863 |
value="Reset", interactive=False, visible=False
|
1864 |
)
|
1865 |
fix_kp_l_info = gr.Markdown(
|
1866 |
+
"""<p style="text-align: center;">② Click 21 keypoints on the image to provide the target hand pose of <b>left hand</b>. See the \"OpenPose keypoints convention\" for guidance.</p>""",
|
1867 |
visible=False
|
1868 |
)
|
1869 |
fix_kp_left = gr.Image(
|
|
|
1894 |
width=LENGTH // 2,
|
1895 |
interactive=False,
|
1896 |
)
|
1897 |
+
|
1898 |
+
# get latent
|
1899 |
with gr.Column():
|
1900 |
gr.Markdown(
|
1901 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">3. Press "Ready" to start pre-processing</p>"""
|
1902 |
)
|
1903 |
fix_ready = gr.Button(value="Ready", interactive=False)
|
1904 |
gr.Markdown(
|
1905 |
+
"""<p style="text-align: center; font-weight: bold; ">Visualized (256, 256)-resized, brushed image</p>"""
|
1906 |
)
|
1907 |
fix_vis_mask32 = gr.Image(
|
1908 |
type="numpy",
|
|
|
1921 |
width=opts.image_size,
|
1922 |
interactive=False,
|
1923 |
)
|
1924 |
+
# gr.Markdown(
|
1925 |
+
# """<p style="text-align: center;">[NOTE] Above should be inpaint mask that you brushed, NOT the segmentation mask of the entire hand. </p>"""
|
1926 |
+
# )
|
1927 |
+
|
1928 |
+
# result column
|
1929 |
with gr.Column():
|
1930 |
gr.Markdown(
|
1931 |
"""<p style="text-align: center; font-size: 20px; font-weight: bold;">4. Press "Run" to get the fixed hand image 🎯</p>"""
|
|
|
1934 |
gr.Markdown(
|
1935 |
"""<p style="text-align: center;">⚠️ >3min and ~24GB per generation</p>"""
|
1936 |
)
|
1937 |
+
fix_result_original = gr.Gallery(
|
1938 |
+
type="numpy",
|
1939 |
+
label="Results on original input",
|
1940 |
+
show_label=True,
|
1941 |
+
height=LENGTH,
|
1942 |
+
min_width=LENGTH,
|
1943 |
+
columns=FIX_MAX_N,
|
1944 |
+
interactive=False,
|
1945 |
+
preview=True,
|
1946 |
+
)
|
1947 |
fix_result = gr.Gallery(
|
1948 |
type="numpy",
|
1949 |
label="Results",
|
|
|
1969 |
)
|
1970 |
fix_clear = gr.ClearButton()
|
1971 |
|
1972 |
+
with gr.Tab("More options"):
|
1973 |
+
gr.Markdown(
|
1974 |
+
"⚠️ Currently, Number of generation > 1 could lead to out-of-memory"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1975 |
)
|
1976 |
+
with gr.Row():
|
1977 |
+
fix_n_generation = gr.Slider(
|
1978 |
+
label="Number of generations",
|
1979 |
+
value=1,
|
1980 |
+
minimum=1,
|
1981 |
+
maximum=FIX_MAX_N,
|
1982 |
+
step=1,
|
1983 |
+
randomize=False,
|
1984 |
+
interactive=True,
|
1985 |
+
)
|
1986 |
+
fix_seed = gr.Slider(
|
1987 |
+
label="Seed",
|
1988 |
+
value=42,
|
1989 |
+
minimum=0,
|
1990 |
+
maximum=10000,
|
1991 |
+
step=1,
|
1992 |
+
randomize=False,
|
1993 |
+
interactive=True,
|
1994 |
+
)
|
1995 |
+
fix_cfg = gr.Slider(
|
1996 |
+
label="Classifier free guidance scale",
|
1997 |
+
value=3.0,
|
1998 |
+
minimum=0.0,
|
1999 |
+
maximum=10.0,
|
2000 |
+
step=0.1,
|
2001 |
+
randomize=False,
|
2002 |
+
interactive=True,
|
2003 |
+
)
|
2004 |
+
fix_quality = gr.Slider(
|
2005 |
+
label="Quality",
|
2006 |
+
value=10,
|
2007 |
+
minimum=1,
|
2008 |
+
maximum=10,
|
2009 |
+
step=1,
|
2010 |
+
randomize=False,
|
2011 |
+
interactive=True,
|
2012 |
+
)
|
2013 |
+
|
2014 |
+
# listeners
|
2015 |
+
# fix_crop.change(resize_to_full, fix_crop, fix_ref)
|
2016 |
+
fix_crop.change(lambda x: x, fix_crop, fix_original) # fix_original: (real_H, real_W, 3)
|
2017 |
+
fix_crop.change(stay_crop, [fix_crop, fix_crop_coord], [fix_crop_coord, fix_ref])
|
2018 |
+
fix_crop.select(process_crop, [fix_crop, fix_crop_coord], [fix_crop_coord, fix_ref])
|
2019 |
+
# fix_ref.change(disable_crop, fix_crop_coord, fix_crop)
|
2020 |
+
fix_ref.change(enable_component, [fix_crop, fix_crop], fix_ref)
|
2021 |
+
fix_ref.change(enable_component, [fix_crop, fix_crop], fix_finish_crop)
|
2022 |
+
fix_finish_crop.click(visualize_ref, [fix_ref], [fix_img])
|
2023 |
+
fix_finish_crop.click(get_mask_inpaint, [fix_ref], [fix_inpaint_mask]) # fix_ref: (real_cropped_H, real_cropped_W, 3)
|
2024 |
fix_img.change(lambda x: x, [fix_img], [fix_kp_right])
|
2025 |
fix_img.change(lambda x: x, [fix_img], [fix_kp_left])
|
2026 |
fix_inpaint_mask.change(
|
|
|
2065 |
],
|
2066 |
)
|
2067 |
fix_kp_right.select(
|
2068 |
+
get_kps, [fix_img, fix_kpts, gr.State("right")], [fix_kp_right, fix_kpts] # fix_img: (real_cropped_H, real_cropped_W, 3)
|
2069 |
)
|
2070 |
fix_undo_right.click(
|
2071 |
undo_kps, [fix_img, fix_kpts, gr.State("right")], [fix_kp_right, fix_kpts]
|
|
|
2087 |
)
|
2088 |
fix_ready.click(
|
2089 |
ready_sample,
|
2090 |
+
[fix_ref, fix_inpaint_mask, fix_kpts],
|
2091 |
[
|
2092 |
fix_ref_cond,
|
2093 |
fix_target_cond,
|
|
|
2106 |
fix_latent,
|
2107 |
fix_inpaint_latent,
|
2108 |
fix_kpts_np,
|
2109 |
+
fix_original,
|
2110 |
+
fix_crop_coord,
|
2111 |
fix_n_generation,
|
2112 |
fix_seed,
|
2113 |
fix_cfg,
|
2114 |
fix_quality,
|
2115 |
],
|
2116 |
+
[fix_result, fix_result_pose, fix_result_original],
|
2117 |
)
|
2118 |
fix_clear.click(
|
2119 |
fix_clear_all,
|
2120 |
[],
|
2121 |
[
|
2122 |
fix_crop,
|
2123 |
+
fix_crop_coord,
|
2124 |
fix_ref,
|
2125 |
+
fix_checkbox,
|
2126 |
fix_kp_right,
|
2127 |
fix_kp_left,
|
2128 |
fix_result,
|
2129 |
fix_result_pose,
|
2130 |
+
fix_result_original,
|
2131 |
fix_inpaint_mask,
|
2132 |
fix_original,
|
2133 |
fix_img,
|
|
|
2145 |
fix_quality,
|
2146 |
],
|
2147 |
)
|
2148 |
+
fix_clear.click(
|
2149 |
+
fix_set_unvisible,
|
2150 |
+
[],
|
2151 |
+
[
|
2152 |
+
fix_kp_right,
|
2153 |
+
fix_kp_left,
|
2154 |
+
fix_kp_r_info,
|
2155 |
+
fix_kp_l_info,
|
2156 |
+
fix_undo_left,
|
2157 |
+
fix_undo_right,
|
2158 |
+
fix_reset_left,
|
2159 |
+
fix_reset_right
|
2160 |
+
]
|
2161 |
+
)
|
2162 |
|
2163 |
gr.Markdown("<h1>Citation</h1>")
|
2164 |
gr.Markdown(
|
brown_logo.png
ADDED
![]() |
Git LFS Details
|
meta_logo.png
ADDED
![]() |
Git LFS Details
|
sbatch/sbatch_demo.sh
ADDED
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/bin/bash
|
2 |
+
|
3 |
+
# job name
|
4 |
+
#SBATCH -J demo_foundhand
|
5 |
+
|
6 |
+
# partition
|
7 |
+
#SBATCH --partition=ssrinath-gcondo --gres=gpu:1 --gres-flags=enforce-binding
|
8 |
+
#SBATCH --account=ssrinath-gcondo
|
9 |
+
|
10 |
+
# ensures all allocated cores are on the same node
|
11 |
+
#SBATCH -N 1
|
12 |
+
|
13 |
+
# cpu cores
|
14 |
+
#SBATCH --ntasks-per-node=4
|
15 |
+
|
16 |
+
# memory per node
|
17 |
+
#SBATCH --mem=32G
|
18 |
+
|
19 |
+
# runtime
|
20 |
+
#SBATCH -t 240:00:00
|
21 |
+
|
22 |
+
# output
|
23 |
+
#SBATCH -o out/demo.out
|
24 |
+
|
25 |
+
# error
|
26 |
+
#SBATCH -e err/demo.err
|
27 |
+
|
28 |
+
# email notifiaction
|
29 |
+
# SBATCH --mail-type=ALL
|
30 |
+
|
31 |
+
module load miniconda3/23.11.0s
|
32 |
+
source /oscar/runtime/software/external/miniconda3/23.11.0/etc/profile.d/conda.sh
|
33 |
+
conda activate handdiff
|
34 |
+
|
35 |
+
cd $HOME/hdd/FoundHand_demo
|
36 |
+
echo Directory is `pwd`
|
37 |
+
|
38 |
+
python -u app.py
|
vqvae.py
CHANGED
@@ -20,7 +20,10 @@ from typing import List
|
|
20 |
import torch
|
21 |
import torch.nn.functional as F
|
22 |
from torch import nn
|
23 |
-
|
|
|
|
|
|
|
24 |
|
25 |
|
26 |
class Autoencoder(nn.Module):
|
|
|
20 |
import torch
|
21 |
import torch.nn.functional as F
|
22 |
from torch import nn
|
23 |
+
try:
|
24 |
+
import spaces
|
25 |
+
except:
|
26 |
+
pass
|
27 |
|
28 |
|
29 |
class Autoencoder(nn.Module):
|