Meta Album Insects Dataset (Micro): 800 Images of Insects in Natural Habitats

The National Museum of Natural History, Paris, and the SPIPOLL citizen science project collected over 290,000 images of insects. This subset contains 800 preprocessed images, each cropped and resized to 128x128 pixels, depicting insects in natural environments like flowers or vegetation. The dataset was curated by Ihsan Ullah for Meta-Album and released in March 2022.

Use Cases

• Few-shot image classification based on the 20 hierarchical insect classes.
• Benchmarking meta-learning algorithms on a standardized, preprocessed image dataset.
• Training models to identify insect orders and families in natural habitats.
• Studying data quality and engagement in citizen science projects using image data.

Strengths

• Contains 800 images, providing a manageable subset for experimentation.
• Images are uniformly preprocessed to 128x128 pixels, ensuring consistent input size.
• Includes 20 hierarchical classes (Order, Super-Family, Family, Texa) for structured classification tasks.
• Originates from a large source collection of over 290,000 images gathered by researchers and volunteers.

Limitations

• Row count for the full dataset is unknown, limiting assessment of overall scale.
• Column-level documentation is absent; field semantics must be inferred after download.
• The CC BY-NC 2.0 license restricts commercial use.

Provenance

Source

SPIPOLL Science project and National Museum of Natural History, Paris.

Collection Method

Images collected by researchers and volunteers via website and mobile apps, then cropped and resized.