Gaspard Dussert
Gaspard Dussert

Research Engineer

About Me

I’m a Research Engineer at the CNRS and working at the Biometry and Evolutionary Biology Lab, UMR 5558). I have a background in applied mathematics, statistics, machine learning and computer vision.

In my PhD, I developed and implemented new methods for the automated analysis of camera trap images, leveraging deep learning to support ecological research. I’m an active member of the DeepFaune project, where I contribute to model training, data curation, and software development.

Outside of research, I’m passionate about wildlife photography and naturalist activities.

Download CV
Interests
  • Artificial Intelligence
  • Computer Vision
  • Ecology
  • Medical Imaging
  • Statistics
Education
  • PhD in Machine Learning

    Université Lyon 1

  • M.Sc. MVA (Mathematics, Vision, Learning)

    ENS Paris-Saclay

  • Engineering Degree in Applied Mathematics

    ENSTA Paris

  • CPGE MPSI - MP*

    Lycée Pierre de Fermat, Toulouse

Experience

  1. Research Engineer (Permanent Position)

    CNRS
    • Development and implementation of methods and tools for data analysis in ecology and evolution.
  2. PhD Student

    University Lyon 1 - LBBE
    • Create new deep learning techniques for automatic species classification in camera trap images: score calibration, behavior prediction and leveraging sequence context
    • Contribute to the improvement and development of the DeepFaune software.
    • PhD supervisors: S. Dray, V. Miele and S. Chamaillé-Jammes
  3. Research Engineer (Fixed-term Contract)

    CNRS - CREATIS
    • Develop a predictive model of consciousness state and 6-month outcomes for coma patients using multiple medical imaging modalities.
    • Utilized advanced AI methods, including transformers and self-supervised learning.
  4. Research Internship

    University Lyon 1 - CREATIS
    • Develop a model for segmentation and characterization of prostate cancer in multiparametric MRI using weak annotations (point-based labels).
    • Investigated non-standard loss functions: modeling inter-class correlations and incorporating constraints on lesion size.

Education

  1. PhD in Machine Learning

    Université Lyon 1
  2. M.Sc. MVA (Mathematics, Vision, Learning)

    ENS Paris-Saclay
  3. Engineering Degree in Applied Mathematics

    ENSTA Paris
  4. CPGE MPSI - MP*

    Lycée Pierre de Fermat, Toulouse
Featured Publications
Recent Publications
(2026). Paying Attention to Other Animal Detections Improves Camera Trap Classification Models. Methods in Ecology and Evolution.
(2025). Camera traps and deep learning enable efficient large-scale density estimation of wildlife in temperate forest ecosystems. Remote Sensing in Ecology and Conservation.
(2025). Zero-shot animal behaviour classification with vision-language foundation models. Methods in Ecology and Evolution.
(2024). Being confident in confidence scores: calibration in deep learning models for camera trap image sequences. Remote Sens Ecol Conserv.
Recent Projects
Recent & Upcoming Talks