ShorterSplatting/full_eval.py at main · MachinePerceptionLab/ShorterSplatting · GitHub

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
#
# 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  george.drettakis@inria.fr
#

"""
python ./full_eval.py --save_images \
    --mipnerf360 xxx \
    --tanksandtemples xxx \
    --deepblending xxx \
    --enable_dash --lambda_entropy 0.015 --scale_reset_factor 0.2
"""


import os
from argparse import ArgumentParser

# 3.3M
MAX_N_GAUSSIAN = {
    "bicycle": 5987095,#54275
    "flowers": 3618411,#38347
    "garden": 5728191,#138766
    "stump": 4867429,#32049
    "treehill": 3770257,#52363
    "room": 1548960,#112627
    "counter": 1190919,#155767
    "kitchen": 1803735,#241367
    "bonsai": 1252367,#206613
    "truck": 2584171,#136029
    "train": 1085480,#182686
    "drjohnson": 3273600,#80861
    "playroom": 2326100#37005
}

mipnerf360_outdoor_scenes = ["bicycle", "flowers", "garden", "stump", "treehill"]
mipnerf360_indoor_scenes = ["room", "counter", "kitchen", "bonsai"]
tanks_and_temples_scenes = ["truck", "train"]
deep_blending_scenes = ["drjohnson", "playroom"]

def build_scene_output_path(output_path, dataset, scene, scale_reset_factor, enable_dash, lambda_entropy):
    path_suffix = f'-{dataset}-litegs'
    if enable_dash:
        path_suffix += '+dash'
    if scale_reset_factor > 0.0:
        path_suffix += f'+reset.{scale_reset_factor}'
    if lambda_entropy > 0:
        path_suffix += f'+entropy.{lambda_entropy}'
    return os.path.join(output_path+path_suffix, scene)

parser = ArgumentParser(description="Full evaluation script parameters")
parser.add_argument("--skip_training", action="store_true")
parser.add_argument("--output_path", default="./output")
parser.add_argument('--mipnerf360', "-m360", required=True, type=str)
parser.add_argument("--tanksandtemples", "-tat", required=True, type=str)
parser.add_argument("--deepblending", "-db", required=True, type=str)
parser.add_argument('--colmap_subfolder', default=".", type=str)
parser.add_argument('--save_images', action="store_true", help="Save rendered and ground truth images")
parser.add_argument('--enable_dash', action="store_true", help="Enable dash mode")
parser.add_argument('--scale_reset_factor', type=float, default=0.0, help="Scale reset factor")
parser.add_argument('--lambda_entropy', type=float, default=0.0, help="Entropy regularization weight")
args, _ = parser.parse_known_args()

all_scenes = []
all_scenes.extend(mipnerf360_outdoor_scenes)
all_scenes.extend(mipnerf360_indoor_scenes)
all_scenes.extend(tanks_and_temples_scenes)
all_scenes.extend(deep_blending_scenes)

if not args.skip_training:
    # Build parameter strings based on command line arguments
    scale_reset_factor_param = f" --scale_reset_factor {args.scale_reset_factor}"
    lambda_entropy_param = f" --lambda_entropy {args.lambda_entropy}"

    for scene in mipnerf360_outdoor_scenes:
        scene_input_path=os.path.join(args.mipnerf360,scene,args.colmap_subfolder)
        scene_output_path=build_scene_output_path(args.output_path, 'm360', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
        test_epochs = " --test_epochs " + " ".join(map(str, range(0, 150)))
        if args.enable_dash:
            final_gaussian_count_param = f' --max_n_gaussian {int(MAX_N_GAUSSIAN[scene])}'
        else:
            final_gaussian_count_param = f' --final_gaussian_count {int(MAX_N_GAUSSIAN[scene])}'
        res = os.system("python example_train.py -s " + scene_input_path + " -i images_4 -m " + scene_output_path + " --eval --sh_degree 3 " + final_gaussian_count_param
                        + scale_reset_factor_param + lambda_entropy_param
                        )
        if res != 0:
            print(f"Training failed for scene {scene}")

    for scene in mipnerf360_indoor_scenes:
        scene_input_path=os.path.join(args.mipnerf360,scene,args.colmap_subfolder)
        scene_output_path=build_scene_output_path(args.output_path, 'm360', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
        test_epochs = " --test_epochs " + " ".join(map(str, range(0, 150)))
        if args.enable_dash:
            final_gaussian_count_param = f' --max_n_gaussian {int(MAX_N_GAUSSIAN[scene])}'
        else:
            final_gaussian_count_param = f' --final_gaussian_count {int(MAX_N_GAUSSIAN[scene])}'
        res = os.system("python example_train.py -s " + scene_input_path + " -i images_2 -m " + scene_output_path + " --eval --sh_degree 3 " + final_gaussian_count_param
                        + scale_reset_factor_param + lambda_entropy_param
                        )
        if res != 0:
            print(f"Training failed for scene {scene}")

    for scene in tanks_and_temples_scenes:
        scene_input_path=os.path.join(args.tanksandtemples,scene,args.colmap_subfolder)
        scene_output_path=build_scene_output_path(args.output_path, 'tat', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
        test_epochs = " --test_epochs " + " ".join(map(str, range(0, 150)))
        if args.enable_dash:
            final_gaussian_count_param = f' --max_n_gaussian {int(MAX_N_GAUSSIAN[scene])}'
        else:
            final_gaussian_count_param = f' --final_gaussian_count {int(MAX_N_GAUSSIAN[scene])}'
        res = os.system("python example_train.py -s " + scene_input_path + " -i images -m " + scene_output_path + " --eval --sh_degree 3 " + final_gaussian_count_param
                        + scale_reset_factor_param + lambda_entropy_param
                        )
        if res != 0:
            print(f"Training failed for scene {scene}")

    for scene in deep_blending_scenes:
        scene_input_path=os.path.join(args.deepblending,scene,args.colmap_subfolder)
        scene_output_path=build_scene_output_path(args.output_path, 'db', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
        test_epochs = " --test_epochs " + " ".join(map(str, range(0, 150)))
        if args.enable_dash:
            final_gaussian_count_param = f' --max_n_gaussian {int(MAX_N_GAUSSIAN[scene])}'
        else:
            final_gaussian_count_param = f' --final_gaussian_count {int(MAX_N_GAUSSIAN[scene])}'
        res = os.system("python example_train.py -s " + scene_input_path + " -i images -m " + scene_output_path + " --eval --sh_degree 3 " + final_gaussian_count_param
                        + scale_reset_factor_param + lambda_entropy_param
                        )
        if res != 0:
            print(f"Training failed for scene {scene}")

save_images_flag = " --save_images" if args.save_images else ""
for scene in mipnerf360_outdoor_scenes:
    scene_input_path=os.path.join(args.mipnerf360,scene,args.colmap_subfolder)
    scene_output_path=build_scene_output_path(args.output_path, 'm360', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
    res = os.system("python example_metrics.py -s " + scene_input_path + " -i images_4 -m " + scene_output_path + " --sh_degree 3" + save_images_flag)
    if res != 0:
        print(f"Evaluation failed for scene {scene}")

for scene in mipnerf360_indoor_scenes:
    scene_input_path=os.path.join(args.mipnerf360,scene,args.colmap_subfolder)
    scene_output_path=build_scene_output_path(args.output_path, 'm360', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
    res = os.system("python example_metrics.py -s " + scene_input_path + " -i images_2 -m " + scene_output_path + " --sh_degree 3" + save_images_flag)
    if res != 0:
        print(f"Evaluation failed for scene {scene}")

for scene in tanks_and_temples_scenes:
    scene_input_path=os.path.join(args.tanksandtemples,scene,args.colmap_subfolder)
    scene_output_path=build_scene_output_path(args.output_path, 'tat', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
    res = os.system("python example_metrics.py -s " + scene_input_path + " -i images -m " + scene_output_path + " --sh_degree 3" + save_images_flag)
    if res != 0:
        print(f"Evaluation failed for scene {scene}")

for scene in deep_blending_scenes:
    scene_input_path=os.path.join(args.deepblending,scene,args.colmap_subfolder)
    scene_output_path=build_scene_output_path(args.output_path, 'db', scene, args.scale_reset_factor, args.enable_dash, args.lambda_entropy)
    res = os.system("python example_metrics.py -s " + scene_input_path + " -i images -m " + scene_output_path + " --sh_degree 3" + save_images_flag)
    if res != 0:
        print(f"Evaluation failed for scene {scene}")