Datasets related problem · Issue #152 · google-research/multinerf · GitHub
Skip to content
This repository was archived by the owner on Feb 11, 2025. It is now read-only.
This repository was archived by the owner on Feb 11, 2025. It is now read-only.

Datasets related problem #152

Open
Open

Description

@chrizz777

What you've done is amazing. Thank you so much for sharing.

I encountered the following problem when running the script train_raw.sh. Where did this transforms.json file come from? Was there a problem with my colmap operation?

image

Activity

  1. sudarshan-iyengar commented on Feb 27, 2024

    @sudarshan-iyengar

    I faced a similar issue as you with no transforms.json file. I had to do a sort of strange workaround, so I'd really appreciate it if someone could give a more straightforward way of doing it, but this is what I did:
    1- downloaded the colmap2nerf.py file from instant-ngp
    2- saved my colmap outputs (cameras.txt, points3D.txt, and images.txt) to a folder called colmap_text
    3- ran the colmap2nerf.py as follows: python colmap2nerf.py --colmap_matcher exhaustive --text path/to/colmap_text --images path/to/images --aabb_scale 16 --overwrite
    4-moved the transforms.json file that was produced by step 3 into the directory on the same level as your images directory
    5- run the training script
    Alternatively:
    Please check to make sure that sparse/0 folder contains the necessary .bin files and isn't empty. If it is empty, there was probably an error with colmap.
    Hope this helps!!

  2. chingguo commented on Feb 29, 2024

    @chingguo

    I faced a similar issue as you with no transforms.json file. I had to do a sort of strange workaround, so I'd really appreciate it if someone could give a more straightforward way of doing it, but this is what I did: 1- downloaded the colmap2nerf.py file from instant-ngp 2- saved my colmap outputs (cameras.txt, points3D.txt, and images.txt) to a folder called colmap_text 3- ran the colmap2nerf.py as follows: python colmap2nerf.py --colmap_matcher exhaustive --text path/to/colmap_text --images path/to/images --aabb_scale 16 --overwrite 4-moved the transforms.json file that was produced by step 3 into the directory on the same level as your images directory 5- run the training script

    Hope this helps!!
    Hello, I have encountered the same problem, but my COLMAP outputs are in ”bin“ format instead of “txt” format. Is there a solution? Thank you very much.

  3. sudarshan-iyengar commented on Mar 5, 2024

    @sudarshan-iyengar

    I would recommend you to check out the colmap documentation, where you can see how to export your model as txt: colmap
    output format

  4. luhan225 commented on Aug 1, 2024

    @luhan225

    I faced a similar issue as you with no transforms.json file. I had to do a sort of strange workaround, so I'd really appreciate it if someone could give a more straightforward way of doing it, but this is what I did: 1- downloaded the colmap2nerf.py file from instant-ngp 2- saved my colmap outputs (cameras.txt, points3D.txt, and images.txt) to a folder called colmap_text 3- ran the colmap2nerf.py as follows: python colmap2nerf.py --colmap_matcher exhaustive --text path/to/colmap_text --images path/to/images --aabb_scale 16 --overwrite 4-moved the transforms.json file that was produced by step 3 into the directory on the same level as your images directory 5- run the training script
    Hope this helps!!
    Hello, I have encountered the same problem, but my COLMAP outputs are in ”bin“ format instead of “txt” format. Is there a solution? Thank you very much.

    Hello, I am currently facing the same problem as you. I would like to ask if you have successfully converted the bin file to a txt file and run it?

  5. gdidbwkos commented on Dec 27, 2024

    @gdidbwkos

    After I do this, there is still no such file or directory: 'transforms.json'

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions