Datasets:

Project-AgML
/

spittlebug_segmentation

dataset_info:
  features:
    - name: image
      dtype: image
    - name: mask
      dtype: image
    - name: split
      dtype: string
  splits:
    - name: train
      num_bytes: 787927393
      num_examples: 347
  download_size: 787977003
  dataset_size: 787927393
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - image-segmentation
size_categories:
  - n<1K

Spittlebug Segmentation

This dataset provides real-world field images of spittlebugs, a common agricultural pest, annotated for semantic segmentation tasks. Images were captured using handheld RGB cameras including Intel RealSense D435, iPhone 11, and Canon EOS1100D during April 2024 and 2025 in Valenzano, Italy. The dataset contains 347 images with pixel-level mask annotations.

This dataset is indexed on https://project-agml.github.io/ as part of the AgML python library.

The original train/test/val split has been preserved in the split column.

Citation

@article{elia2026sustainable,
  title={Towards sustainable management of Xylella fastidiosa vectors: An annotated image dataset for automated in-field detection of Aphrophoridae foam},
  author={Elia, Michele and Cardellicchio, Angelo and Paradiso, Michele and Veronico, Giuseppe and Rana, Arianna and Petitti, Antonio and Renò, Vito and Pascuzzi, Simone and Milella, Annalisa},
  journal={Data in Brief},
  volume={65},
  pages={112477},
  year={2026},
  publisher={Elsevier}
}

This dataset was reformatted from its original format to match HuggingFace standards.