Plant identification: Experts vs. machines in the era of deep learning: deep learning techniques challenge flora experts

Bonnet Pierre, Goeau Hervé, Hang Siang Thye, Lasseck Mario, Sulc Milan, Malécot Valery, Jauzein Philippe, Melet Jean-Claude, You Christian, Joly Alexis. 2018. Plant identification: Experts vs. machines in the era of deep learning: deep learning techniques challenge flora experts. In : Multimedia tools and applications for environmental and biodiversity informatics.. Joly Alexis (ed.), Vrochidis Stefanos (ed.), Karatzas Kostas (ed.), Karppinen Ari (ed.), Bonnet Pierre (ed.). Cham : Springer, pp. 131-149. (Multimedia Systems and Applications) ISBN 978-3-319-76444-3

Book chapter
[img] Published version - Anglais
Access restricted to CIRAD agents
Use under authorization by the author or CIRAD.

Télécharger (7MB) | Request a copy

Abstract : Automated identification of plants and animals have improved considerably in the last few years, in particular thanks to the recent advances in deep learning. The next big question is how far such automated systems are from the human expertise. Indeed, even the best experts are sometimes confused and/or disagree between each others when validating visual or audio observations of living organism. A picture or a sound actually contains only a partial information that is usually not sufficient to determine the right species with certainty. Quantifying this uncertainty and comparing it to the performance of automated systems is of high interest for both computer scientists and expert naturalists. This chapter reports an experimental study following this idea in the plant domain. In total, nine deeplearning systems implemented by three different research teams were evaluated with regard to nine expert botanists of the French flora. Therefore, we created a small set of plant observations that were identified in the field and revised by experts in order to have a near-perfect golden standard. The main outcome of this work is that the performance of state-of-the-art deep learning models is now close to the most advanced human expertise. This shows that automated plant identification systems are now mature enough for several routine tasks, and can offer very promising tools for autonomous ecological surveillance systems.

Mots-clés Agrovoc : Informatique, Identification, Flore, Surveillance de l’environnement, Automatisation, Botaniste

Mots-clés géographiques Agrovoc : France

Classification Agris : C30 - Documentation and information
F70 - Plant taxonomy and geography
F40 - Plant ecology

Champ stratégique Cirad : Axe 6 (2014-2018) - Sociétés, natures et territoires

Auteurs et affiliations

  • Bonnet Pierre, CIRAD-BIOS-UMR AMAP (FRA) ORCID: 0000-0002-2828-4389
  • Goeau Hervé, CIRAD-BIOS-UMR AMAP (FRA)
  • Hang Siang Thye, Toyohashi University of Technology (JPN)
  • Lasseck Mario, Museum Für Naturkunde Berlin (DEU)
  • Sulc Milan, Czech Technical University in Prague (CZE)
  • Malécot Valery, Agrocampus Ouest (FRA)
  • Jauzein Philippe, AgroParisTech (FRA)
  • Melet Jean-Claude
  • You Christian, Societe botanique centre ouest (FRA)
  • Joly Alexis, INRIA (FRA)

Autres liens de la publication

Source : Cirad-Agritrop (

View Item (staff only) View Item (staff only)

[ Page générée et mise en cache le 2021-02-02 ]