Bandeau traitement de données

Use of multiple object tracking and subject re-identification to elucidate group dynamics and interactions among Insects

PhD research project awarded in the 2023 call for projects.

  • Title: "Use of multiple object tracking and subject re-identification to elucidate group dynamics and interactions among insects" ; INRAE & Inria.
  • PhD student: Yuan Gao.
  • Affiliated unit: INRAE SophiaAgroBioTech.
  • Co-supervisors:
    • Vincent Calcagno, INRAE SophiaAgroBioTech.
    • François Bremond, Inria, Sophia Antipolis.
  • Doctoral school: STIC, Université Paris-Saclay.

Project summary:

Animals are individual agents acting within groups or populations. Despite the common assumption that all individuals are essentially replicas of average behavior, it is increasingly clear that individuals often exhibit substantial differences ("personalities") even though they share the same genome. Furthermore, interactions between individuals, especially agonistic interactions, dynamically alter the expression of this variation. Understanding and quantifying these interindividual variations are crucial for better predicting the functioning of population groups. This is particularly true for microscopic insects used as biocontrol agents, which are released into crops in very high densities, and whose interactions directly determine pest suppression effectiveness. However, the study of the behavior of large groups has remained an unresolved challenge due to the inability to maintain individual identities over time. New artificial intelligence techniques involving Multiple Object Tracking (MOT) and Subject Re-identification (ReID) can now overcome this limitation. Using the biocontrol agent Trichogramma as a model system, this thesis will employ innovative MOT-ReID techniques to study interindividual behavioral differences within groups of insects and to investigate the impact of these differences on parasitism and biocontrol efficiency.

See also