TQTQliu / MVSGaussian

[ECCV 2024] MVSGaussian: Fast Generalizable Gaussian Splatting Reconstruction from Multi-View Stereo
https://mvsgaussian.github.io/
MIT License
393 stars 19 forks source link

Error when running run.py #58

Open ZhenyuSun-Walker opened 2 weeks ago

ZhenyuSun-Walker commented 2 weeks ago

Hello sir, happy Nation Day!I've met some difficulties running the run.py.

Workspace:  .
configs/mvsgs/dtu_pretrain.yaml
configs/mvsgs/colmap_eval.yaml
EXP NAME:  dtu_pretrain
load model: ./trained_model/mvsgs/dtu_pretrain/latest.pth
Traceback (most recent call last):
  File "run.py", line 98, in <module>
    globals()['run_' + args.type]()
  File "run.py", line 55, in run_evaluate
    data_loader = make_data_loader(cfg, is_train=False)
  File "/home/sunzhenyu/Projects/MVSGaussian/lib/datasets/make_dataset.py", line 90, in make_data_loader
    dataset = make_dataset(cfg, is_train)
  File "/home/sunzhenyu/Projects/MVSGaussian/lib/datasets/make_dataset.py", line 39, in make_dataset
    dataset = dataset(**args)
  File "lib/datasets/colmap/mvsgs.py", line 19, in __init__
    self.build_metas()
  File "lib/datasets/colmap/mvsgs.py", line 51, in build_metas
    c2ws = c2ws[train_ids]
IndexError: index 4 is out of bounds for axis 0 with size 4

However, the colmap step is nothing wrong. I don't know which part went wrong, so I'm going to ask for your constructive advice.

TQTQliu commented 2 weeks ago

Thanks! This means that train_ids is out of bounds of c2ws. Please check the value of train_ids and the shape of c2ws.

crazyzyz commented 2 days ago

Hello sir, happy Nation Day!I've met some difficulties running the run.py.

Workspace:  .
configs/mvsgs/dtu_pretrain.yaml
configs/mvsgs/colmap_eval.yaml
EXP NAME:  dtu_pretrain
load model: ./trained_model/mvsgs/dtu_pretrain/latest.pth
Traceback (most recent call last):
  File "run.py", line 98, in <module>
    globals()['run_' + args.type]()
  File "run.py", line 55, in run_evaluate
    data_loader = make_data_loader(cfg, is_train=False)
  File "/home/sunzhenyu/Projects/MVSGaussian/lib/datasets/make_dataset.py", line 90, in make_data_loader
    dataset = make_dataset(cfg, is_train)
  File "/home/sunzhenyu/Projects/MVSGaussian/lib/datasets/make_dataset.py", line 39, in make_dataset
    dataset = dataset(**args)
  File "lib/datasets/colmap/mvsgs.py", line 19, in __init__
    self.build_metas()
  File "lib/datasets/colmap/mvsgs.py", line 51, in build_metas
    c2ws = c2ws[train_ids]
IndexError: index 4 is out of bounds for axis 0 with size 4

However, the colmap step is nothing wrong. I don't know which part went wrong, so I'm going to ask for your constructive advice.

Hey bro! I hope to know how you finally dealt with it, because I also encountered this problem, your answer is very important to me, looking forward to your reply