princeton-computational-imaging / neural-point-light-fields Goto Github PK

Hi,
I run the script python main.py --config example_configs/config_kitti_0006_example_render.txt --render_only --manipulate dance
I met the error below.

Traceback (most recent call last):
  File "<string>", line 2, in <module>
  File "/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/util/dispatch.py", line 180, in wrapper
    return target(*args, **kwargs)
  File "/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/ops/array_ops.py", line 3956, in gather
    params, indices, axis, name=name)
  File "/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/ops/gen_array_ops.py", line 4075, in gather_v2
    _six.raise_from(_core._status_to_exception(e.code, message), None)
  File "<string>", line 3, in raise_from
tensorflow.python.framework.errors_impl.InvalidArgumentError: indices[1] = -1 is not in [0, 11) [Op:GatherV2]

I can run the scrpt python main.py --config example_configs/config_kitti_0006_example_render.txt --render_only --manipulate switch_location successful.

Thank you for releasing the code. I wonder if you can kindly double-check and fix some bugs in the code. I am running into various errors when trying to replicate your code. I would appreciate it if you can kindly make it error-free.
Thank you.

Some errors that I encountered:

1. neural-point-light-fields/src/datasets/preprocess_Waymo.py

   Line 106 in 6996a11

   (range_images, camera_projections, range_image_top_pose) = \

   
   
   There are also multiple open3d errors.

2. neural-point-light-fields/src/scenes/__init__.py

   Line 140 in 6996a11

   point_chunk_size=exp_dict.get('point_chunk', 1e12),

   .
   

Hi,

Thanks for the source code you provided! There is an issue in the trainval.py when you want to use multi GPUs for training. You have this part:

`jc = None
if os.path.exists("job_config.py"):
import job_config

    jc = job_config.JOB_CONFIG
    if args.ngpus > 1:
        jc["resources"]["gpu"] = args.ngpus
        python_binary_path += (
            f" -m torch.distributed.launch --nproc_per_node={args.ngpus} --use_env "
        )`

which never works since there is no "job_config.py" in the repo, could you please fix this issue?

Thanks for the great project!

I would like to follow your experimental setting on waymo. However, I am a little confused about the chosen scenes for experiments. In the paper, it is stated that 5 static scenes from waymo are chosen for evaluation.

However, in the supplementary, there are 6 scenes in total. I wonder which 5 scenes are chosen from the 6 scenes exactly.

Looking forward to the reply and thanks in advance!

Hi, I tried following the url provided in the readme to download the validation files but it shows me an empty bucket in google cloud storage where I cannot see the filelist, and hence cannot download the files. Can you please let me know if there is any other way to access these files in the format required for Neural Point Light Fields?

I noticed that the data is also available at the Waymo Open Dataset website, under the dataset names 'motion' and 'perception'. Perhaps you can tell me which version of which dataset is compatible with this repository and I can try to download from the source directly.

Thanks a lot!

Thanks a lot for sharing your code. May I ask what is the license type (e.g., MIT, BSD, Apache, etc.) for using the code?

-- Nhon

Seems that the compressed Waymo dataset does not work