Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

bundle splits with dataset #27

Closed
Tracked by #20
AndreiBarsan opened this issue Apr 23, 2023 · 4 comments · Fixed by #35
Closed
Tracked by #20

bundle splits with dataset #27

AndreiBarsan opened this issue Apr 23, 2023 · 4 comments · Fixed by #35
Assignees

Comments

@AndreiBarsan
Copy link
Contributor

AndreiBarsan commented Apr 23, 2023

This just means that there is a method to load the partitions from the devkit. This could be done with text files in the repo itself, or putting them in the AWS bucket.

@AndreiBarsan AndreiBarsan mentioned this issue Apr 23, 2023
9 tasks
@una-dinosauria
Copy link
Contributor

Exporting the partitions as text takes 400+ MB, while in npz they take ~30 MB, so I think I'll put them in the bucket

@una-dinosauria
Copy link
Contributor

I've put them under s3://pit30m/partitions/, I'll make a PR to actually use them in a dataset

@AndreiBarsan
Copy link
Contributor Author

Fantastic, thank you!

AndreiBarsan pushed a commit that referenced this issue May 19, 2023
Bundles partition into the dataloader. The dataloader may now receive an
additional argument `partitions`, and provides a boolean array that
indicates which sensor measurements (currently, from the front camera),
should be skipped. If `partitions` is `None`, then no frames are skipped
at all.

Closes #27
@una-dinosauria
Copy link
Contributor

I've replaced the partitions under s3://pit30m/partitions/ with the ones computed from dense utm poses, as discussed in https://github.com/pit30m/pit30m-internal/issues/11

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants