-
Notifications
You must be signed in to change notification settings - Fork 8
/
demo.py
264 lines (223 loc) · 12.3 KB
/
demo.py
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
import os
import pprint
import random
import numpy as np
import torch
import torch.nn.parallel
import torch.optim
import torch.nn.functional as F
import itertools
import torch.utils.data
import torch.utils.data.distributed
import torch.distributed as dist
import argparse
from config.config import config, update_config
from utils import exp_utils, train_utils, eval_utils, vis_utils, geo_utils
from models.model import FORGE as ReconModel
from models.model import sequence_from_distance, chose_selected
from models.perceptual_loss import VGGPerceptualLoss as PerceptualLoss
from dataset.kubric import Kubric
import time
from pytorch3d.renderer import look_at_view_transform
import json
from PIL import Image
def run_demo(args, config, dataset, data, model, model_gt, output_dir, device):
# data in [b,t,c,h,w]
model.eval()
model_gt.eval()
_, t, c, h, w = data.shape
b = 1
for batch_idx, cur_data in enumerate(data):
clips = cur_data.float().unsqueeze(0).to(device) # [b=1,t,c,h,w]
K_cv2 = torch.tensor([[1.38888, 0.0, 0.5], [0.0, 1.38888, 0.5], [0.0, 0.0, 0.0]]) * 256.0
K_cv2[-1,-1] = 1.0
K_cv2 = K_cv2.unsqueeze(0).float()
# predict 3D feature volumes
clips = clips.reshape(b*t,c,h,w)
features_raw = model.module.encoder_3d.get_feat3D(clips) # [b*t,C,D,H,W]
_, C, D, H, W = features_raw.shape
clips = clips.reshape(b,t,c,h,w)
features_raw = features_raw.reshape(b,t,C,D,H,W)
# predict camera relative pose
pose_feat_3d = model.module.encoder_traj(features_raw, return_features=True) # [b*(t-1),1024]
pose_feat_2d = model.module.encoder_traj_2d(clips, return_features=True) # [b*(t-1),1024]
pose_feat = torch.cat([pose_feat_3d, pose_feat_2d], dim=-1) # [b*(t-1),2048]
pred = model.module.pose_head(pose_feat) # [b*(t-1), 8]
poses_cam, _ = pred.split([model.module.encoder_traj.pose_dim, 1], dim=-1)
tmp = torch.zeros_like(poses_cam)
tmp[:,:4] = F.normalize(poses_cam[:,:4])
tmp[:,4:] = poses_cam[:,4:]
poses_cam = tmp
# optimize pose
poses_cam = refine_pose(model, dataset, clips, poses_cam.detach(), features_raw.detach(), K_cv2.unsqueeze(0), device)
# get camera extrinsics and pose
camPoseRel_cv2 = model.module.encoder_traj.toSE3(poses_cam)
canonical_pose_cv2 = dataset.get_canonical_pose_cv2(device=device) # [4,4]
canonical_extrinsics_cv2 = dataset.get_canonical_extrinsics_cv2(device=device)
camPoses_cv2 = canonical_pose_cv2.unsqueeze(0) @ camPoseRel_cv2
camE_cv2 = torch.inverse(camPoses_cv2) # [b*(t-1),4,4], canonicalized extrinsics
camE_cv2 = camE_cv2.reshape(b,t-1,4,4)
camPoses_cv2 = camPoses_cv2.reshape(b,t-1,4,4)
camPoses_cv2 = torch.cat([canonical_pose_cv2.reshape(1,1,4,4).repeat(b,1,1,1), camPoses_cv2], dim=1)# [b,t,4,4]
camE_cv2 = torch.cat([canonical_extrinsics_cv2.reshape(1,1,4,4).repeat(b,1,1,1), camE_cv2], dim=1) # [b,t,4,4]
# feature fusion and predict neural volume
features_transformed = model_gt.module.rotate(voxels=features_raw, camPoses_cv2=camPoses_cv2[:,:t], grid_size=D) # [b,t,C=128,D=16,H,W]
idxs = sequence_from_distance(camPoses_cv2[:,:,:3,3])
features_transformed = chose_selected(features_transformed, idxs)
features_mv = model_gt.module.encoder_3d.fuse(features_transformed) # [b,t,C,D,H,W] -> [b,C,D,H,W]
densities_mv = model_gt.module.encoder_3d.get_density3D(features_mv) # [b,1,D,H,W]
features_mv = model_gt.module.encoder_3d.get_render_features(features_mv) # [b,C,D,H,W]
# sample NVS camera poses
num_views_all = 4 * 7
elev = torch.linspace(0, 0, num_views_all)
azim = torch.linspace(0, 360, num_views_all) + 180
NVS_R_all, NVS_T_all = look_at_view_transform(dist=config.render.camera_z, elev=elev, azim=azim)
NVS_pose_all = torch.cat([NVS_R_all, NVS_T_all.view(-1,3,1)], dim=-1)
# 360-degree NVS
for idx in range(b):
rendered_imgs_results, rendered_masks_results = [], []
all_feature = features_mv[idx].unsqueeze(0).repeat(7,1,1,1,1) # [N,C,D,H,W]
all_density = densities_mv[idx].unsqueeze(0).repeat(7,1,1,1,1).clamp(max=1.0) # [N,1,D,H,W]
for pose_idx in range(4):
cameras = {
'K': K_cv2.repeat(7,1,1), # [N,3,3]
'R': NVS_pose_all[pose_idx*7: (pose_idx+1)*7,:3,:3],
'T': NVS_pose_all[pose_idx*7: (pose_idx+1)*7,:3,3],
}
rendered_imgs, rendered_masks = model_gt.module.render(cameras, all_feature, all_density) # [N,c,h,w], [N,1,h,w]
rendered_imgs_results.append(rendered_imgs.detach())
rendered_masks_results.append(rendered_masks.detach())
rendered_imgs_results = torch.cat(rendered_imgs_results, dim=0)
rendered_masks_results = torch.cat(rendered_masks_results, dim=0)
print('Saving results to {}'.format(output_dir))
vis_utils.vis_NVS(imgs=rendered_imgs_results,
masks=rendered_masks_results,
img_name=str(batch_idx) + '_' + str(idx),
output_dir=output_dir,
subfolder='vis_360')
def refine_pose(model, dataset, clips, poses_cam, features, K, device):
b, t, c, h, w = clips.shape
_, _, C, D, H, W = features.shape
masks = (clips[:, :, 0:1] > 0.05).float()
target_imgs = clips.view(b*t,c,h,w)
target_masks = masks.view(b*t,1,h,w)
# optimize poses
poses_cam = poses_cam.detach()
poses_cam.requires_grad = True
poses_cam_rot = poses_cam[:,:4].detach()
poses_cam_trans = poses_cam[:,4:].detach()
poses_cam_rot.requires_grad = True
poses_cam_trans.requires_grad = True
#optimizer = torch.optim.SGD([poses_cam], lr=0.0005, momentum=0.9)
optimizer = torch.optim.Adam([{'params': poses_cam_rot, 'lr': 0.001},
{'params': poses_cam_trans, 'lr': 0.0005}
], lr=0.001)#, momentum=0.9)
for iter_idx in range(2000+1): # 500 iterations should be enough
poses_cam_normalized = torch.zeros_like(poses_cam)
poses_cam_normalized[:,:4] = F.normalize(poses_cam_rot)
poses_cam_normalized[:,4:] = poses_cam_trans
camPoseRel_cv2 = model.module.encoder_traj.toSE3(poses_cam_normalized)
# get camera extrinsics and pose for rendering
canonical_pose_cv2 = dataset.get_canonical_pose_cv2(device=device) # [4,4]
canonical_extrinsics_cv2 = dataset.get_canonical_extrinsics_cv2(device=device)
camPoses_cv2 = canonical_pose_cv2.unsqueeze(0) @ camPoseRel_cv2
camE_cv2 = torch.inverse(camPoses_cv2) # [b*(t-1),4,4], canonicalized extrinsics
camE_cv2 = camE_cv2.reshape(b,t-1,4,4)
camPoses_cv2 = camPoses_cv2.reshape(b,t-1,4,4)
camPoses_cv2 = torch.cat([canonical_pose_cv2.reshape(1,1,4,4).repeat(b,1,1,1), camPoses_cv2], dim=1)
camE_cv2 = torch.cat([canonical_extrinsics_cv2.reshape(1,1,4,4).repeat(b,1,1,1), camE_cv2], dim=1)
# transform features
features_transformed = model.module.rotate(voxels=features, camPoses_cv2=camPoses_cv2[:,:t], grid_size=D) # [b,t,C,D,H,W]
idxs = sequence_from_distance(camPoses_cv2[:,:,:3,3])
features_transformed = chose_selected(features_transformed, idxs)
features_mv = model.module.encoder_3d.fuse(features_transformed) # [b,t,C=128,D=16,H,W] -> [b,C,D,H,W]
densities_mv = model.module.encoder_3d.get_density3D(features_mv) # [b,1,D=32,H,W]
features_mv = model.module.encoder_3d.get_render_features(features_mv) # [b,C=16,D=32,H,W]
_, C2, D2, H2, W2 = features_mv.shape
# render
camE_cv2 = camE_cv2.repeat(1,1,1,1) # [b,2*t,4,4]
camPoses_cv2 = camPoses_cv2.repeat(1,1,1,1) # [b,2*t,4,4]
camK = K.repeat(1,t,1,1) # [b,2*t,3,3]
cameras = {
'R': camE_cv2.reshape(b*1*t,4,4)[:,:3,:3].to(device), # [b*t,3,3]
'T': camE_cv2.reshape(b*1*t,4,4)[:,:3,3].to(device), # [b*t,3]
'K': camK.reshape(b*1*t,3,3).to(device) # [b*t,3,3]
}
features_all = features_mv.unsqueeze(1).repeat(1,t,1,1,1,1).reshape(b*t,C2,D2,H2,W2) # [b,2*t,C,D,H,W] -> [b*2*t,C,D,H,W]
densities_all = densities_mv.unsqueeze(1).repeat(1,t,1,1,1,1).reshape(b*t,1,D2,H2,W2)
rendered_imgs, rendered_masks = model.module.render(cameras, features_all, densities_all, return_origin_proj=False)
# calculate loss
loss_recon_img = config.loss.recon_rgb * F.mse_loss(rendered_imgs, target_imgs)
loss_recon_mask = config.loss.recon_mask * F.mse_loss(rendered_masks, target_masks)
loss_recon = loss_recon_img + loss_recon_mask
# optimize pose
optimizer.zero_grad()
loss_recon.backward()
optimizer.step()
poses_final = torch.zeros_like(poses_cam)
poses_final[:,:4] = F.normalize(poses_cam_rot)
poses_final[:,4:] = poses_cam_trans
return poses_final
def parse_args():
parser = argparse.ArgumentParser(description='FORGE demo')
parser.add_argument(
'--cfg', help='experiment configure file name', required=True, type=str)
args, rest = parser.parse_known_args()
update_config(args.cfg)
return args
def main():
# Get args and config
args = parse_args()
logger, output_dir, tb_log_dir = exp_utils.create_logger(config, args.cfg, phase='train')
print('=> Saving args and config into logger...')
logger.info(pprint.pformat(args))
logger.info(pprint.pformat(config))
# set random seeds
torch.cuda.manual_seed_all(config.seed)
torch.manual_seed(config.seed)
np.random.seed(config.seed)
random.seed(config.seed)
# set device
gpus = range(torch.cuda.device_count())
device = torch.device('cuda') if len(gpus) > 0 else torch.device('cpu')
# get model
model = ReconModel(config).to(device)
cpt_root = './output/kubric/joint_pose_2d3d/pred_pose_2d3d_joint'
cpt_name = 'cpt_best_psnr_26.340881009038913_7.545314707482719.pth.tar'
cpt = torch.load(os.path.join(cpt_root, cpt_name))['state_dict']
model.load_state_dict(cpt, strict=True)
model = torch.nn.DataParallel(model)
# use fusion module without joint training
model_gt = ReconModel(config).to(device)
cpt_root = './output/kubric/gt_pose/gt_pose'
cpt_name = 'cpt_best_psnr_31.842686198427398.pth.tar'
cpt = torch.load(os.path.join(cpt_root, cpt_name))['state_dict']
del cpt['encoder_traj.out.3.weight']
del cpt['encoder_traj.out.3.bias']
model_gt.load_state_dict(cpt, strict=False)
model_gt = torch.nn.DataParallel(model_gt)
all_data = []
for i in range(3): # case id
cur_data = []
for j in range(3): # frame id
img_name = './assets/real_images/{}_{}'.format(i+1, j)
if os.path.isfile(img_name+'.jpg'):
with Image.open(img_name+'.jpg') as img_pil:
img_np = np.asarray(img_pil)[:,:,:3]
elif os.path.isfile(img_name+'.png'):
with Image.open(img_name+'.png') as img_pil:
img_np = np.asarray(img_pil)[:,:,:3].copy()
mask_np = np.uint8(np.asarray(img_pil)[:,:,3:] > 0).copy()
img_np *= mask_np
rgb = Image.fromarray(img_np[:,:,:3])
rgb = rgb.resize((256, 256), Image.ANTIALIAS)
rgb = np.asarray(rgb).transpose((2,0,1)) / 255.0 # [3,H,W]
rgb = torch.from_numpy(rgb)
cur_data.append(rgb)
cur_data = torch.stack(cur_data) # [t,3,h,w]
all_data.append(cur_data)
all_data = torch.stack(all_data) # [b,t,3,h,w]
val_data = Kubric(config, split='test')
run_demo(args, config, dataset=val_data, data=all_data, model=model, model_gt=model_gt, output_dir=output_dir, device=device)
if __name__ == '__main__':
main()