Swedish Wildflowers Dataset with 90 Species and 13,496 Images

Swedish native flora is represented by 13,496 images across 90 wildflower species. The dataset was created by Renuka Devi Chappidi and is designed for machine learning image classification tasks. Images were collected from iNaturalist, GBIF, and Wikimedia Commons.

Use Cases

• Train image classifiers to identify Swedish wildflower species based on the 90 species labels.
• Benchmark computer vision models on a geographically specific flora dataset.
• Conduct transfer learning experiments for plant recognition tasks.
• Analyze model performance across a dataset split into 10,796 training, 1,349 validation, and 1,351 test images.

Strengths

• Contains 13,496 images, providing a substantial volume for model training.
• Covers 90 distinct Swedish wildflower species, offering a specific taxonomic scope.
• Includes predefined splits for training (10,796 images), validation (1,349), and testing (1,351).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect geographic and source bias inherent to the collection platforms (iNaturalist, GBIF, Wikimedia Commons).

Provenance

Source

iNaturalist, GBIF, Wikimedia Commons

Collection Method

Collected and curated by Renuka Devi Chappidi.

Time Range

null

Freshness

Last updated 2026-03-29 20:49:44; freshness should be verified.

Geography

Sweden