General information
Offer title : Post-Doctoral Student [M/F]: Design of visual functions for robotics (H/F)
Reference : UPR8001-PATDAN-009
Number of position : 1
Workplace : TOULOUSE
Date of publication : 29 October 2024
Type of Contract : Researcher in FTC
Contract Period : 24 months
Expected date of employment : 3 February 2025
Proportion of work : Full Time
Remuneration : Monthly gross salary is between 2991€ and 3417€ depending on experience.
Desired level of education : Doctorate
Experience required : Indifferent
Section(s) CN : 7 - Information sciences: processing, integrated hardware-software systems, robots, commands, images, content, interactions, signals and languages
Missions
This position is open in the Robotics, Action and Perception team (RAP, https://www.laas.fr/en/teams/rap/) of LAAS-CNRS, in Toulouse, France. The research of the RAP team concerns the design, implementation and evaluation of algorithms for visual perception and perception-movement coupling in dynamic environments. The recruited postdoctoral researcher will work in close collaboration with the permanent staff and doctoral students of the team on current and future projects. His/Her research work in a robotics context will focus on:
o Computer vision (2D and/or 3D approaches to detection, tracking, target recognition).
o Machine learning (supervised or unsupervised, semi-supervised).
o The (homogeneous or heterogeneous) fusion of exteroceptive sensor data.
Activities
The tasks of the recruited person will consist of the design and prototyping of visual functions in the emblematic case studies of the research team: (1) human-robot interaction by analysis of facial expressions, recognition of posture, gestures or human actions; (2) mobility (perception of heterogeneous users, navigation, mapping…); (3) agricultural robotics (modeling and segmentation of evolving plants, semantic segmentation…). These functions will then be validated on benchmark datasets made available by the “computer vision”, “automotive” or “agricultural robotics” communities. The aim will be to integrate them on robotic platforms (mobile robots, mobile manipulators) in order to evaluate and validate them experimentally. The missions of the recruited postdoctoral researcher will be refined and modulated according to his/her scientific profile.
Skills
This offer is specifically aimed at a holder of a PhD in vision for robotics. The expected skills for this position are:
o Very good knowledge of Vision for Robotics and Machine Learning.
o Mastery of open-source libraries: Pytorch or equivalent, OpenCV, Open3D, PCL.
o Programming languages: C, C++, Python.
o Practical experience on robotic platforms will be appreciated. Experience on architectures dedicated to the embedding of visual functions will be an additional asset.
o Autonomy, teamwork.
o Writing skills in French and English.
Work Context
This work will take place in the Robotics Action Perception team, within the Robotics Department of LAAS-CNRS, Toulouse, France.
The gross monthly salary is between €2,991 and €3,417 depending on experience.
The position is located in a sector covered by the protection of scientific and technical potential (PPST), and therefore requires, in accordance with the regulations, that recruitment be authorized by the competent authority of the MESR.
The position is located in a sector under the protection of scientific and technical potential (PPST), and therefore requires, in accordance with the regulations, that your arrival is authorized by the competent authority of the MESR.
Constraints and risks
N/A
Additional Information
Candidates must put together a file that clearly establishes their qualifications and motivation for this position. The “Curriculum Vitae” file to be submitted to the emploi.cnrs.fr website must contain the following elements:
- Candidate's resume, including a list of significant publications related to the requested profile.
- Cover letter.
- Pre-defense and defense reports of thesis.
- Transcript of grades from the final year of studies.
- Possibly, letters of recommendation.