Adapting a global plant identification model to detect invasive alien plant species in high-resolution road side images

Vincent Espitalier, Jean-Christophe Lombardo, Hervé Goëau, Christophe Botella, Toke Thomas Høye, Mads Dyrmann, Pierre Bonnet, Alexis Joly

Abstract: Highlights•Developed models to detect invasive plant species in high-resolution roadside images.•Adapted a global model for single-plant identification, pre-trained on over six million images.•Explored two complementary approaches: tiling and resolution-adaptive processing for detecting both single and multiple species per image.•Tiling improves generalization to roadside scenes without extra fine-tuning.•Results demonstrate effective detection while reducing annotation costs.

External IDs:dblp:journals/ecoi/EspitalierLGBHDBJ25