[RFC]A potential solution to address the issue #9394 by by enlarging grid to (H+1, W+1) - seeking feedback

[zy1git]

Address this issue #9394 by enlarging grid to (H+1, W+1).

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/9403

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 2 Unrelated Failures

As of commit 73067ed with merge base 4967c64 ():

NEW FAILURES - The following jobs have failed:

• Build Windows Wheels / pytorch/vision / upload / upload-wheel-py3_10-cuda12_6 (gh)
  Unable to download artifact(s): Artifact not found for name: pytorch_vision__3.10_cu126_x64
• Build Windows Wheels / pytorch/vision / upload / upload-wheel-py3_10-cuda12_8 (gh)
  Unable to download artifact(s): Artifact not found for name: pytorch_vision__3.10_cu128_x64

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• Build Windows Wheels / pytorch/vision / build-wheel-py3_10-cuda12_6 (gh) (trunk failure)
  Process completed with exit code 1.
• Build Windows Wheels / pytorch/vision / build-wheel-py3_10-cuda12_8 (gh) (trunk failure)
  Process completed with exit code 1.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[NicolasHug]

Thanks for trying this, Can you share visual outputs and compare with the alternative of claming to H-1, W-1?
I think it might help figuring out what should be the most correct approach

[zy1git]

@NicolasHug I prepared a script and get the visual outputs but I feel that the visual outputs are quite similar because the difference is around 1-2 pixels which our human eyes cannot tell. However, I can print the bbox values and you will notice the difference.