remove clamp_keypoints from transforms

[AntoineSimoulin]

Context

This PR proposes an alternative to #9234. We disable the clamping for KeyPoints after the following transformations vertical_flip_keypoints, horizontal_flip_keypoints, _affine_keypoints_with_expand, pad_keypoints, crop_keypoints, perspective_keypoints, and elastic_keypoints. We are fixing this to avoid the issue observed in #9223, where clamping modifies the position of KeyPoints and create a misalignment with the actual locations in the transformed image. This PR will need to be supplemented with #9235 to implement a class similar to SanitizeBoundingBoxes transform to remove non valid KeyPoints, e.g. outside of the image canvas.

Implementation details

In the current proposition, we simply do not apply the clamp_keypoints by default after any transform. For KeyPoints, the clamping must therefore be explicitly set with the ClampKeyPoints transform.

# we manually set bounding boxes and key points for the demo; in a real pipeline, these would come from the data or a CV model’s output.
orig_pts = tv_tensors.KeyPoints(
    [
        [
            [445, 700],  # nose
            [320, 660],
            [370, 660],
            [420, 660],  # left eye
            [300, 620],
            [420, 620],  # left eyebrow
            [475, 665],
            [515, 665],
            [555, 655],  # right eye
            [460, 625],
            [560, 600],  # right eyebrow
            [370, 780],
            [450, 760],
            [540, 780],
            [450, 820],  # mouth
        ],
    ],
    canvas_size=(orig_img.size[1], orig_img.size[0]),
    clamping_mode="hard"
)
cropper = v2.Compose(
    [
        v2.RandomCrop(size=(128, 128)),
        # v2.ClampKeyPoints(),  # must set to enable clamping
    ]
)
crops = [cropper((orig_img, orig_pts)) for _ in range(4)]
plot([(orig_img, orig_pts)] + crops)

Illustration of the changes

Without explicit clamping

With explicit clamping (default before)

Testing

Please run the following unit tests

pytest test/test_transforms_v2.py -vvv -k "keypoints"
...
661 passed, 9057 deselected in 4.16s

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/9236

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• Driver update on H100 and A100 instances

✅ No Failures

As of commit 859622b with merge base 7a13ad0 ():
💚 Looks good so far! There are no failures yet. 💚

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

[NicolasHug]

Thank you @AntoineSimoulin !

[github-actions]

Hey @AntoineSimoulin!

You merged this PR, but no labels were added.
The list of valid labels is available at https://github.com/pytorch/vision/blob/main/.github/process_commit.py