-
Notifications
You must be signed in to change notification settings - Fork 29
/
config.py
49 lines (43 loc) · 1.79 KB
/
config.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
import time
class Config:
# training settings
train_image_width = 256
train_image_height = 256
train_min_depth = 0.25
train_max_depth = 20.0
train_n_depth_levels = 64
train_minimum_pose_distance = 0.125
train_maximum_pose_distance = 0.325
train_crawl_step = 3
train_subsequence_length = None
train_predict_two_way = None
train_freeze_batch_normalization = False
train_data_pipeline_workers = 8
train_epochs = 100000
train_print_frequency = 5000
train_validate = True
train_seed = int(round(time.time()))
# test settings
test_image_width = 320
test_image_height = 256
test_distortion_crop = 0
test_perform_crop = False
test_visualize = True
test_n_measurement_frames = 2
test_keyframe_buffer_size = 30
test_keyframe_pose_distance = 0.1
test_optimal_t_measure = 0.15
test_optimal_R_measure = 0.0
# SET THESE: TRAINING FOLDER LOCATIONS
dataset = "/media/ardaduz/T5/train"
train_run_directory = "/home/ardaduz/Workspace/git/deep-video-mvs/training-runs"
# SET THESE: TESTING FOLDER LOCATIONS
# for run-testing-online.py (evaluate a single scene, WITHOUT keyframe indices, online selection)
test_online_scene_path = "/home/ardaduz/Workspace/git/deep-video-mvs/sample-data/hololens-dataset/000"
# for run-testing.py (evaluate all available scenes, WITH pre-calculated keyframe indices)
test_offline_data_path = "/home/ardaduz/Workspace/git/deep-video-mvs/sample-data"
# below give a dataset name like tumrgbd, i.e. folder or None
# if None, all datasets will be evaluated given that
# their keyframe index files are in Config.test_offline_data_path/indices folder
test_dataset_name = "hololens-dataset" # or None
test_result_folder = "/media/ardaduz/T5/results/"