Support empty label fields

[sklum]

For my use case, it's valuable to be able to train on images without any labels of any kind as a part of a batch. This change allows that.

Summary by Sourcery

Enhancements:

• Handle empty label fields during data processing.

[sourcery-ai]

Reviewer's Guide by Sourcery

This pull request modifies the _process_label_fields method in albumentations/core/utils.py to support training on images without any labels. It adds a check to handle empty encoded_labels gracefully, preventing errors during concatenation with data_array.

Sequence diagram for processing label fields with empty labels

sequenceDiagram
  participant A as Albumentations
  participant L as LabelManager
  A->>A: _process_label_fields(data, data_name)
  loop for each label_field in label_fields
    A->>L: process_field(data_name, label_field, data[label_field])
    L-->>A: encoded_labels (empty)
    alt encoded_labels.size > 0
      A->>A: np.hstack((data_array, encoded_labels))
    else encoded_labels.size == 0
      A->>A: data_array (no change)
    end
  end
  A-->>A: return data_array

Loading

File-Level Changes

Change	Details	Files
Modified the _process_label_fields method to handle empty label fields gracefully.	Added a conditional check to ensure encoded_labels has a size greater than zero before concatenating it with data_array. If encoded_labels is empty, data_array is initialized as an empty numpy array to avoid errors during concatenation.	albumentations/core/utils.py

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!
• Generate a plan of action for an issue: Comment @sourcery-ai plan on
  an issue to generate a plan of action for it.

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey @sklum - I've reviewed your changes - here's some feedback:

Overall Comments:

• Consider adding a test case that covers the scenario with empty labels.

Here's what I looked at during the review

• 🟢 General issues: all looks good
• 🟢 Security: all looks good
• 🟢 Testing: all looks good
• 🟢 Complexity: all looks good
• 🟢 Documentation: all looks good

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[ternaus]

@sklum

For my use case, it's valuable to be able to train on images without any labels of any kind as a part of a batch. This change allows that.

What do you mean?

You can pass images, and no any other labels to nearly every transform.

[sklum]

If e.g. the bounding boxes are an empty tensor (i.e. bboxes = tensor([])), the call to hstack will fail with the following error:

ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 2 dimension(s)

[sklum]

Here's an MRE which should hopefully clarify my use case. Is this not idiomatic?

import albumentations as A
import numpy as np
from torch import tensor


def main():
    img_size = 640

    train_transforms = A.Compose(
        [
            A.LongestMaxSize(max_size=img_size, interpolation=1),
            A.PadIfNeeded(min_height=img_size, min_width=img_size),
        ],
        bbox_params=A.BboxParams(
            format="coco",
            label_fields=["labels", "ids"],
            clip=True,
        ),
    )

    image = np.random.rand(640, 640, 3)

    target = {
        "bbox": tensor(
            [
                [174.2637, 116.5894, 264.3941, 315.8969],
                [1.8412, 40.0401, 602.9775, 462.5326],
            ]
        ),
        "labels": tensor([1, 2]),
    }

    # Works fine
    train_transforms(
        image=image,
        bboxes=target["bbox"],
        labels=target["labels"].tolist(),
        ids=range(len(target["labels"].tolist())),
    )

    # Image has no classes that I'm interested in detecting
    target = {
        "bbox": tensor([]),
        "labels": tensor([]),
    }

    """
    File "/Users/scottklum/Developer/albumentations_mre/.venv/lib/python3.10/site-packages/albumentations/core/utils.py", line 212, in _process_label_fields
    data_array = np.hstack((data_array, encoded_labels))
    File "/Users/scottklum/Developer/albumentations_mre/.venv/lib/python3.10/site-packages/numpy/_core/shape_base.py", line 365, in hstack
        return _nx.concatenate(arrs, 0, dtype=dtype, casting=casting)
    ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 2 dimensio
    """
    train_transforms(
        image=image,
        bboxes=target["bbox"],
        labels=target["labels"].tolist(),
        ids=range(len(target["labels"].tolist())),
    )


if __name__ == "__main__":
    main()