Pl@ntNet app in the era of deep learning

Antoine Affouard, Hervé Goeau, Pierre Bonnet, Jean-Christophe Lombardo, Alexis Joly

Abstract: Pl@ntNet is a large-scale participatory platform and information system dedicated to the production of botanical data through image-based plant identification. In June 2015, Pl@ntNet mobile front-ends moved from classical hand-crafted visual features to deep-learning based image representations. This paper gives an overview of today's Pl@ntNet architecture and discusses how the introduction of convolutional neural networks did improve the whole workflow along the years.

TL;DR: We synthetise the deep-learning based architecture of Pl@ntNet application and its societal impact

Conflicts: inria.fr

Keywords: Computer vision, Supervised Learning, Applications