RandomRotate/RandomFlip break image_coord alignment during training
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 52/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Quiet
- Tech stack
- python
- Domain
- computer-vision, machine-learning
Research direction
Start with Dataset.get_data() and the image configuration in configs/nuscenes/semseg-pt-v3m1-0-image.py, then inspect how RandomRotate and RandomFlip transform point-cloud coord. Trace whether precomputed image_coord remains aligned through the training pipeline; done means the transforms preserve correspondence between coord and image_coord for image-based training samples.
Written by the indexing model from the issue text.
Description
Thanks for the great work! I've been using pt-v3m1-0-image and noticed a potential training-time alignment issue:
image configs (e.g. configs/nuscenes/semseg-pt-v3m1-0-image.py) apply RandomRotate and RandomFlip on point cloud, which only modify coord without updating image_coord.
image_coord is pre-computed in Dataset.get_data() before the transform pipeline.
this should silently break the alignment.
If this is true, a large fraction of training samples carry misaligned image features, which hurts classes relying on precise spatial correspondence.
- Dominant language
- Python
- Stars
- 113
- Forks
- 8
- PR merge metrics
- No merged PRs in 30d
Getting set up
We have not checked this project's setup files yet. Start from its README, and see our first-contribution guide for the general steps.
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from VisualComputingInstitute/ditr
-
Difficulty 3/5 1-2 days Newbie friendliness 48/100
-
Difficulty 5/5 Over a week Newbie friendliness 25/100
VisualComputingInstitute/ditr#13 · 1 comment ·
-
Inference demoOpen
Difficulty 4/5 3-5 days Newbie friendliness 35/100
VisualComputingInstitute/ditr#12 · 7 comments ·
-
Difficulty 4/5 3-5 days Newbie friendliness 45/100
VisualComputingInstitute/ditr#11 · 2 comments ·
All issues in VisualComputingInstitute/ditr
Similar issues
-
adr
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
kristofdegrave/homeassistant-smart-charging#1607 ·
Maintainers usually reply within 1 day
-
namespace operations
Difficulty 2/5 1-3 hours Newbie friendliness 64/100
EclipseFdn/open-vsx.org#13665 ·
Maintainers usually reply within 1 day
-
doc good first issue help wanted
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
collective/icalendar#1865 · 2 comments ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
canonical/opentelemetry-collector-operator#409 ·
Maintainers usually reply within 1 day
-
Difficulty 1/5 Under an hour Newbie friendliness 85/100
mozilla/addons-release-tests#1243 ·
Maintainers usually reply within 1 day