Postdoctorat dans le projet ANR 4DPlants

Université de Strasbourg; France; Hybride; EUR 40 000 - 50 000; Plein temps

Résumé du poste

A prestigious university in France is seeking a postdoctoral researcher for the ANR 4DPlants project, focusing on the semantic and instance segmentation of 3D point clouds of growing plants. Candidates must have a PhD in computer science, particularly in computer vision or machine learning, along with strong programming skills in C++ and Python. The position offers a 12-month contract with opportunities for remote work. Ideal candidates should demonstrate team communication abilities and a strong research orientation.

Qualifications

• PhD in computer science with specialization in computer vision, digital geometry processing, or machine learning.
• Strong programming skills in C++ and Python are essential.

Responsabilités

• Design and implement a space-time registration method for 3D point clouds.
• Evaluate the registration method qualitatively and quantitatively.
• Communicate results through presentations and publications.

Connaissances

C++ programming

Python programming

Algorithm design

Team communication

Formation

PhD in computer science

Outils

Git

Overleaf

Jupyter notebooks

Description du poste

Organisation/Company Université de Strasbourg Department Direction des ressources humaines Research Field Computer science Researcher Profile Recognised Researcher (R2) Positions Postdoc Positions Country France Application Deadline 9 Jan 2026 - 23:59 (Europe/Paris) Type of Contract Temporary Job Status Full-time Hours Per Week 37h30 Offer Starting Date 1 Feb 2026 Is the job funded through the EU Research Framework Programme? Not funded by a EU programme Is the Job related to staff position within a Research Infrastructure? No

Offer Description

Position identification

• Title of post : Postdoctoral position in the ANR 4DPlants project
• Type of contract : postdoc
• Category (A,B or C) : A
• Contract/project period : 12 months
• Expected date of employment : 01/02/2026
• Proportion of work : 100 %
• Workplace : ICube research unit, IGG team (University of Strasbourg)
• Desired level of education : PhD in computer science
• Experience required : PhD in computer science, with publications in international journals and/or in proceedings of international conferences

Contact(s) for information on the position (identity, position, e-mail address, telephone) : Prof. Franck Hétroy‑Wheeler, 4DPlants project leader, [e-mail]

Date of publication : 12/12/2025

Closing date for the receipt of applications : 09/01/2026

Research project or operation

The subject of this postdoctoral position falls within the scope of the ANR project 4DPlants (https://4dplants.icube.unistra.fr/ ) funded by the French Ministry of Higher Education and Research, which involves partners from ICube/Université de Strasbourg, RDP/ENS Lyon, and Inria Grenoble. The overall objective of the 4DPlants project is to develop new methods for the semantic and instance segmentation of time-varying photogrammetry 3D point clouds of growing plants for high throughput phenotyping applications. A key part of the project is to craft training data for a Deep Learning-based method aimed at predicting procedural plant representations. Those training data should be obtained from real datasets with annotated organs at each time step. The high‑level of expertise and time required to produce such annotations make a manual approach intractable. This postdoctoral position focuses on providing a space‑time registration of a growing plant. Specifically, given a temporal sequence of 3D point clouds {P1, P2, …, PN} of a growing plant and supposing semantic and instance segmentations of the plant organs is given at least for P1, we want to automatically propagate consistent segmentation labels through the whole point cloud sequence and find deformations for each organ. The method must be robust to fine branching structures and organ events, including the appearance of new organs, that lead to new semantic or instance labels, as well as the change or loss of some other ones (e.g., buds opening, or leaf senescence). To ensure scalability, the method should be efficient in terms of time and memory consumption, and as automatic as possible.

Activities

• Bibliography related to the project and writing of a state‑of‑the‑art report.
• Design and implementation of a space‑time registration method for 3D point clouds of growing plants.
• Thorough qualitative and quantitative evaluation of the method.
• Communication of the results through conference presentations and publication in an international journal.

• Participation in the IGG research team activities, including to the regular meetings.
• Presentations of the work to workshops and conferences, publications.

Skills

Qualifications/knowledge :

• PhD in computer science, with a specialisation in computer vision, digital geometry processing and/or machine learning. No specific knowledge about plants is required.

Operational skills/expertise :

• Algorithmic and programming skills in C++ and Python.
• Experience in using Git, Overleaf, Jupyter notebooks.

Personal qualities :

• Ability to communicate and to work in a team.

Presentation of the laboratory/unity : See https://igg.icube.unistra.fr/en/index.php/Main_Page

Hierarchical relationship :

• The postdoc will work within the 4DPlants project, in close collaboration with Prof. Franck Hétroy‑Wheeler (PI), Dr Rémi Allègre and Dr Joris Ravaglia.

Special conditions of practice : Work can partially be done remote, under conditions.