-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathrender.py
executable file
·212 lines (179 loc) · 11.6 KB
/
render.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
#
# Copyright (C) 2023, Inria
# GRAPHDECO research group, https://team.inria.fr/graphdeco
# All rights reserved.
#
# This software is free for non-commercial, research and evaluation use
# under the terms of the LICENSE.md file.
#
# For inquiries contact [email protected]
#
import torch
from scene import Scene
import os
from tqdm import tqdm
from os import makedirs
from gaussian_renderer import render
import torchvision
from utils.general_utils import safe_state
from argparse import ArgumentParser
from arguments import ModelParams, PipelineParams, get_combined_args
from gaussian_renderer import GaussianModel
import sklearn
import sklearn.decomposition
import numpy as np
from PIL import Image
import torch.nn as nn
from models.networks import CNN_decoder, CNN_scale_decoder
from scene.dataset_readers import read_sam_clip_feature
import matplotlib.pyplot as plt
import glob
def feature_visualize_saving(feature):
fmap = feature[None, :, :, :] # torch.Size([1, 512, h, w])
fmap = nn.functional.normalize(fmap, dim=1) # channel维度上做归一化
pca = sklearn.decomposition.PCA(3, random_state=42) # PCA降维到3维
f_samples = fmap.permute(0, 2, 3, 1).reshape(-1, fmap.shape[1])[::3].cpu().numpy() # 1, h, w, 512 -> h*w/3, 512
transformed = pca.fit_transform(f_samples)
feature_pca_mean = torch.tensor(f_samples.mean(0)).float().cuda()
feature_pca_components = torch.tensor(pca.components_).float().cuda()
q1, q99 = np.percentile(transformed, [1, 99])
feature_pca_postprocess_sub = q1
feature_pca_postprocess_div = (q99 - q1)
del f_samples
vis_feature = (fmap.permute(0, 2, 3, 1).reshape(-1, fmap.shape[1]) - feature_pca_mean[None, :]) @ feature_pca_components.T
vis_feature = (vis_feature - feature_pca_postprocess_sub) / feature_pca_postprocess_div
vis_feature = vis_feature.clamp(0.0, 1.0).float().reshape((fmap.shape[2], fmap.shape[3], 3)).cpu() # h, w, 3
return vis_feature
def scale_visualize_saving(scale_map):
max_scale=torch.argmax(scale_map,dim=0) # 0,1,2
max_scale=max_scale/2
return max_scale
def process_scale_map(scale_map):
scale_maps = [torch.zeros_like(scale_map).cuda() for _ in range(3)]
for i, sm in enumerate(scale_maps):
sm[i] = 1
return scale_maps
def process_feature_map(view, scale_map):
gt_feature_maps = []
for sm in process_scale_map(scale_map):
gt_feature_map, mask = read_sam_clip_feature(view.img_embed.cuda(), view.seg_map.cuda(), sm.cuda(), max_mode=True)
gt_feature_maps.append(gt_feature_map * mask)
return gt_feature_maps
def render_set(model_path, source_path, name, iteration, views, gaussians, pipeline, background, speedup, feature_mode, feature_npy, render_mode):
image_gt_list = glob.glob(os.path.join(source_path, 'images', '*.*'))
image_gt_list.sort()
orig_img_width, orig_img_height = Image.open(image_gt_list[0]).size
print("gt image size:", orig_img_height, "," , orig_img_width)
device0 = 'cuda'
device1 = 'cpu'
render_path = os.path.join(model_path, name, "ours_{}".format(iteration), "renders")
# gts_path = os.path.join(model_path, name, "ours_{}".format(iteration), "gt")
feature_map_path = os.path.join(model_path, name, "ours_{}".format(iteration), "feature_map")
gt_feature_map_path = os.path.join(model_path, name, "ours_{}".format(iteration), "gt_feature_map")
scale_map_path=os.path.join(model_path, name, "ours_{}".format(iteration), "scale_map")
feature_map_npy_path = os.path.join(model_path, name, "ours_{}".format(iteration), "feature_map_npy")
#encoder_ckpt_path = os.path.join(model_path, "encoder_chkpnt{}.pth".format(iteration))
decoder_ckpt_path = os.path.join(model_path, "decoder_chkpnt{}.pth".format(iteration))
scale_decoder_ckpt_path=os.path.join(model_path, "scale_decoder_chkpnt{}.pth".format(iteration))
depth_data_path=os.path.join(model_path, name, "ours_{}".format(iteration), "depths")
depth_visual_path=os.path.join(model_path, name, "ours_{}".format(iteration), "depths_visual")
if feature_mode and feature_npy==False:
feature_out_dim=views[0].img_embed.shape[1] # 512 / 768
feature_in_dim = int(feature_out_dim/32)
if speedup:
cnn_decoder = CNN_decoder(feature_in_dim, feature_out_dim)
cnn_decoder.load_state_dict(torch.load(decoder_ckpt_path)['module_state_dict'])
cnn_scale_decoder=CNN_scale_decoder(feature_in_dim, 3)
cnn_scale_decoder.load_state_dict(torch.load(scale_decoder_ckpt_path)['module_state_dict'])
if feature_npy:
makedirs(feature_map_npy_path, exist_ok=True)
elif feature_mode==False:
makedirs(render_path, exist_ok=True)
# makedirs(gts_path, exist_ok=True)
if render_mode == "RGB+ED":
makedirs(depth_data_path, exist_ok=True)
makedirs(depth_visual_path, exist_ok=True)
else:
makedirs(feature_map_path, exist_ok=True)
makedirs(gt_feature_map_path, exist_ok=True)
makedirs(scale_map_path, exist_ok=True)
for idx, view in enumerate(tqdm(views, desc="Rendering progress")):
if feature_npy or feature_mode==False: # change to full resolution
view.image_width = orig_img_width
view.image_height = orig_img_height
render_pkg = render(view, gaussians, pipeline, background, feature_mode=feature_mode, render_mode=render_mode)
feature_map = render_pkg["render"] # 16,731,989
if feature_npy: # only save *.npy feature map file
np.save(os.path.join(feature_map_npy_path, '{0:05d}'.format(idx) + ".npy"), feature_map.permute(1,2,0).cpu().numpy()) # H, W, 16
elif feature_mode==False: # rendering RGB(&depth)
c,h,w=feature_map.shape
# gt = view.original_image[0:3, :, :]
rendering_RGB=feature_map[:3, :, :] # 3，H，W
if c == 4: # including depth
rendering_depth=feature_map[3, :, :] # H，W
rendering_depth=rendering_depth.cpu().numpy()
# save depth
np.save(os.path.join(depth_data_path, view.image_name + "_depth.npy"),rendering_depth)
# save visualized depth
plt.figure("Image",(w/100.0,h/100.0),dpi=100) # 图像窗口名称
plt.imshow(rendering_depth, cmap='viridis')
plt.axis('off')
plt.subplots_adjust(left=0, right=1, bottom=0, top=1)
plt.savefig(os.path.join(depth_visual_path, view.image_name + "_depth.png"), bbox_inches='tight', pad_inches=0)
plt.close()
# save render RGB & gt RGB (use image name)
torchvision.utils.save_image(rendering_RGB, os.path.join(render_path, view.image_name + "_RGB.png"))
# torchvision.utils.save_image(gt, os.path.join(gts_path, view.image_name + ".png"))
else: # rendering feature map
scale_map=cnn_scale_decoder(feature_map.detach()) # 3, 731, 989
torchvision.utils.save_image(scale_map, os.path.join(scale_map_path, '{0:05d}'.format(idx) + ".png"))
scale_map_save=scale_visualize_saving(scale_map)
torchvision.utils.save_image(scale_map_save, os.path.join(scale_map_path, '{0:05d}'.format(idx) + "_class.png"))
gt_feature_map, mask = read_sam_clip_feature(view.img_embed.cuda(),view.seg_map.cuda(),scale_map.cuda(),max_mode=True) # 512, 47, 64
gt_feature_map = gt_feature_map * mask
gt_feature_map_s, gt_feature_map_m, gt_feature_map_l = process_feature_map(view, scale_map)
# torchvision.utils.save_image(render_pkg["render"], os.path.join(render_path, '{0:05d}'.format(idx) + ".png"))
# torchvision.utils.save_image(gt, os.path.join(gts_path, '{0:05d}'.format(idx) + ".png"))
# torchvision.utils.save_image(mask.to(torch.float32), os.path.join(gt_feature_map_path, '{0:05d}'.format(idx) + "_mask.png"))
if speedup:
feature_map = cnn_decoder(feature_map) # C, H, W
feature_map_vis = feature_visualize_saving(feature_map)
Image.fromarray((feature_map_vis.cpu().numpy() * 255).astype(np.uint8)).save(os.path.join(feature_map_path, '{0:05d}'.format(idx) + "_feature_vis.png"))
gt_feature_map_vis = feature_visualize_saving(gt_feature_map)
Image.fromarray((gt_feature_map_vis.cpu().numpy() * 255).astype(np.uint8)).save(os.path.join(gt_feature_map_path, '{0:05d}'.format(idx) + "_feature_vis.png"))
gt_feature_map_s_vis=feature_visualize_saving(gt_feature_map_s)
Image.fromarray((gt_feature_map_s_vis.cpu().numpy() * 255).astype(np.uint8)).save(os.path.join(gt_feature_map_path, '{0:05d}'.format(idx) + "_feature_vis_s.png"))
gt_feature_map_m_vis=feature_visualize_saving(gt_feature_map_m)
Image.fromarray((gt_feature_map_m_vis.cpu().numpy() * 255).astype(np.uint8)).save(os.path.join(gt_feature_map_path, '{0:05d}'.format(idx) + "_feature_vis_m.png"))
gt_feature_map_l_vis=feature_visualize_saving(gt_feature_map_l)
Image.fromarray((gt_feature_map_l_vis.cpu().numpy() * 255).astype(np.uint8)).save(os.path.join(gt_feature_map_path, '{0:05d}'.format(idx) + "_feature_vis_l.png"))
def render_sets(dataset : ModelParams, iteration : int, pipeline : PipelineParams, skip_train : bool, skip_test : bool, feature_mode: bool, feature_npy: bool, render_mode: str): ###
with torch.no_grad():
gaussians = GaussianModel(dataset.sh_degree)
scene = Scene(dataset, gaussians, load_iteration=iteration, shuffle=False) # 从.ply文件中加载场景信息
bg_color = [1,1,1] if dataset.white_background else [0, 0, 0]
background = torch.tensor(bg_color, dtype=torch.float32, device="cuda")
dataset.speedup = getattr(dataset, 'speedup', False)
if not skip_train:
render_set(dataset.model_path, dataset.source_path, "train", scene.loaded_iter, scene.getTrainCameras(), gaussians, pipeline, background, dataset.speedup, feature_mode, feature_npy, render_mode)
if not skip_test:
render_set(dataset.model_path, dataset.source_path, "test", scene.loaded_iter, scene.getTestCameras(), gaussians, pipeline, background, dataset.speedup, feature_mode, feature_npy, render_mode)
if __name__ == "__main__":
# Set up command line argument parser
parser = ArgumentParser(description="Testing script parameters")
model = ModelParams(parser, sentinel=True)
pipeline = PipelineParams(parser)
parser.add_argument("--iteration", default=-1, type=int)
parser.add_argument("--skip_train", action="store_true")
parser.add_argument("--skip_test", action="store_true")
parser.add_argument("--quiet", action="store_true")
parser.add_argument('--feature_mode', action='store_true', help='use feature replace RGB')
parser.add_argument("--feature_npy", action='store_true', help='store 16-dim feature map in npy')
parser.add_argument("--render_mode", default="RGB", type=str) # RGB+ED
args = get_combined_args(parser) #从命令行获取参数，并根据得到的model_path解析该目录下cfg_args文件，获取训练参数
print("Rendering " + args.model_path)
assert not (args.feature_mode and args.render_mode == "RGB+ED"), "Feature mode does not support depth rendering"
# Initialize system state (RNG)
safe_state(args.quiet)
print(args)
render_sets(model.extract(args), args.iteration, pipeline.extract(args), args.skip_train, args.skip_test, args.feature_mode, args.feature_npy, args.render_mode) ###cnn