TypeError: readColmapSceneInfo() missing 2 required positional arguments: 'eval' and 'train_test_exp'

[andrewsilver1997]

Hi,

I was trying to run your code with mipnerf360 dataset. Then I had:

Traceback (most recent call last):
  File "train_kmeans.py", line 409, in <module>
    training(lp.extract(args), op.extract(args), pp.extract(args), args.test_iterations, args.save_iterations,
  File "train_kmeans.py", line 49, in training
    scene = Scene(dataset, gaussians)
  File "/home/zikai/Desktop/compact3d/gaussian-splatting/scene/__init__.py", line 49, in __init__
    scene_info = sceneLoadTypeCallbacks["Colmap"](args.source_path, args.images, args.eval)
TypeError: readColmapSceneInfo() missing 2 required positional arguments: 'eval' and 'train_test_exp'

This is my run.sh file:

path_base=dataset
ncls=4096
ncls_sh=512
ncls_dc=4096
kmeans_iters=10
st_iter=15000
max_iters=30000
max_prune_iter=20000
lambda_reg=1e-7
output_base=output/exp_001
dset=360_v2
scene=bicycle
cuda_device=0
port=4060
ckpt=output/exp_001_noquant/"$dset"/"$scene"/chkpnt"$st_iter".pth
path_source="$path_base"/"$dset"/"$scene"
path_output="$output_base"/"$dset"/"$scene"

CUDA_VISIBLE_DEVICES=$cuda_device python train_kmeans.py \
  --port $port \
  -s="$path_source" \
  -m="$path_output" \
  --start_checkpoint "$ckpt" \
  --kmeans_ncls "$ncls" \
  --kmeans_ncls_sh "$ncls_sh" \
  --kmeans_ncls_dc "$ncls_dc" \
  --kmeans_st_iter "$st_iter" \
  --kmeans_iters "$kmeans_iters" \
  --total_iterations "$max_iters" \
  --quant_params sh dc rot scale\
  --kmeans_freq 100 \
  --opacity_reg \
  --lambda_reg "$lambda_reg" \
  --max_prune_iter "$max_prune_iter" \
  --eval

thank you for your help ;)

[arghavan-kpm]

Hi, thanks for your interest in our work! This error is due to the update in gaussian_splatting code base. def readColmapSceneInfo has been changed in scene/dataset_readers.py with two new arguments args.depths and args.train_test_exp. You may resolve this discrepancy by adding these two arguments and pass them via sceneLoadTypeCallbacks in scene/__init__.py", line 49.

[andrewsilver1997]

thank you! now the problem is solved ;)