-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathrun_step1.sh
54 lines (44 loc) · 2.49 KB
/
run_step1.sh
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
# step1: distill to meshneus
# train init neus with dtu mask
#python train_init_nerf.py -O2 --workspace ./res/dtu_scan105-neus-mask \
#--lambda_eikonal 0.01 --batch_rays 512 --if_mask True \
#--backbone neus --train_resolution_level 1 --eval_resolution_level 4 \
#--data_type 'dtu' --data_path ./data/dtu_scan105 --H 1200 --W 1600 --eval_interval 100 --bound 1.0
# train background instant-ngp
python train_init_nerf.py -O --workspace ./res/dtu_scan105-grid \
--batch_rays 1024 --iters 30000 \
--backbone grid --train_resolution_level 1 --eval_resolution_level 4 \
--data_type 'dtu' --data_path ./data/dtu_scan105 --H 1200 --W 1600 --eval_interval 20 --bound 1.0
# train init neus
python train_init_nerf.py -O2 --workspace ./res/dtu_scan105-neus \
--lambda_eikonal 0.1 --batch_rays 512 \
--backbone neus --train_resolution_level 1 --eval_resolution_level 4 \
--data_type 'dtu' --data_path ./data/dtu_scan105 --H 1200 --W 1600 --eval_interval 100 --bound 1.0
# train diss
python train_distill.py --backbone mesh_neus \
--prior_mesh ./data/dtu_scan105/doll.ply --iters 45000 \
--train_resolution_level 1 --eval_resolution_level 4 --w 1600 --h 1200 -O2 \
--far_point_thresh 0.05 --lambda_eikonal 0.01 \
--lambda_diss 1. \
--teacher_path res/dtu_scan105-neus/checkpoints/df_ep4600.pth \
--sample_R_path data/dtu_scan105/Orient_R.npy \
--R_path '' \
--fine_sigma_net True --fine_ln True \
--prenet_path ./pre_mesh_net/pre_color_net.pth \
--radius_range 0.45 0.55 --fovy_range 50 50 --pose_sample_strategy '360' --phi_range -120 120 --theta_range 30 75 \
--data_type 'dtu' --data_path ./data/dtu_scan105 --H 300 --W 400 \
--num_steps 64 --upsample_steps 64 \
--workspace ./res/diss-dtu_scan105-doll --eval_interval 100 --bound 1.
python train_distill.py --backbone mesh_neus \
--prior_mesh ./data/dtu_scan105/doll.ply --iters 300000 \
--train_resolution_level 1 --eval_resolution_level 4 --w 1600 --h 1200 -O2 \
--far_point_thresh 0.05 --lambda_eikonal 0.01 \
--lambda_diss 1. \
--teacher_path res/dtu_scan105-neus/checkpoints/df_ep4600.pth \
--sample_R_path data/dtu_scan105/Orient_R.npy \
--R_path '' \
--prenet_path ./pre_mesh_net/pre_color_net.pth \
--radius_range 0.45 0.55 --fovy_range 50 50 --pose_sample_strategy '360' --phi_range -120 120 --theta_range 30 75 \
--data_type 'dtu' --data_path ./data/dtu_scan105 --H 300 --W 400 \
--num_steps 64 --upsample_steps 64 \
--workspace ./res/diss-dtu_scan105-doll --eval_interval 100 --bound 1.