Portuguese citizen science and social media images of "Cortaderia Selloana"

This data/code was developed and used in the context of the research article: <em>Can citizen science and social media images support the early detection of new invasion sites? A deep learning test case with Cortaderia selloana;</em> Ana Sofia Cardoso, Eva Malta-Pinto, Siham Tabik, Tom August, Helen E. Roy, Ricardo Correia, Joana R. Vicente, Ana Sofia Vaz; Ecological Informatics; doi. In the first dataset (Citizen_science_dataset), we compiled 2,273 images with <em>Cortaderia selloana</em> from the main Portuguese citizen science platform on invasive alien plants (https://invasoras.pt/) and iNaturalist (https://www.inaturalist.org/), from which only those containing the species with plumes (1,323) were retained. This set of 1,323 images was manually checked, and low resolution and blurred images were removed (resulting in 1,237 images). A similar number (1,237) of outdoor images without <em>Cortaderia selloana</em> (in similar context to the ones with <em>Cortaderia selloana</em>) was additionally collected from Flickr, resulting in a total of 2,474 images. The second dataset (Social_media_dataset) consists of publicly available images taken in Portugal from four social media platforms: Facebook (https://www.facebook.com/), Flickr (https://www.flickr.com/), Instagram (https://www.instagram.com/) and Twitter (https://twitter.com/). Image collection was done by searching for common names of <em>C. selloana</em> in Portuguese (e.g., “Erva das pampas”, “Capim das pampas”), as well as the scientific name of the species (i.e., “<em>Cortaderia selloana</em>”), resulting in a total of 400 images. We also collected the same number of outdoor images without the <em>C. selloana</em> from the same social media platforms). The final size of the transferability set was 800 images.