PhD Position F/M Open PhD grant available at Inria Grenoble on “AI-driven safe motion planning & driving decision-making for autonomous driving”
Inria
Montbonnot-Saint-Martin, 38330
PhD Position F/M Open PhD grant available at Inria Grenoble on “AI-driven safe motion planning & driving decision-making for autonomous driving”
Le descriptif de l’offre ci-dessous est en Anglais
Type de contrat : CDD
Contrat renouvelable : Oui
Niveau de diplôme exigé : Bac + 5 ou équivalent
Fonction : Doctorant
A propos du centre ou de la direction fonctionnelle
The Centre Inria de l’Université de Grenoble groups together almost 600 people in 22 research teams and 7 research support departments.
Staff is present on three campuses in Grenoble, in close collaboration with other research and higher education institutions (Université Grenoble Alpes, CNRS, CEA, INRAE, …), but also with key economic players in the area.
The Centre Inria de l’Université Grenoble Alpe is active in the fields of high-performance computing, verification and embedded systems, modeling of the environment at multiple levels, and data science and artificial intelligence. The center is a top-level scientific institute with an extensive network of international collaborations in Europe and the rest of the world.
Contexte et atouts du poste
Mission confiée
The autonomous driving task involves four main interconnected components: environment perception, tracking, motion planning & driving decision-making. Motion planning & driving decision-making are challenging and critical tasks that involve performing safe navigation in intricate and dense traffic scenarios when subjected to several uncertainties from perceptive sensors such as occlusions and limited field of view. Additionally, the intentions of surrounding traffic participants are also uncertain, making the task a complex problem.
Motion planning can be categorized in two types based on the effect of the planned trajectories on the predicted dynamic agents in the environment: Uni-directional planning, where the ego-vehicle planning considers the predictions of the dynamic agents, while those predictions are independent of the planned future trajectories of the ego-vehicle [1]. And bi-directional planning, where a two-way association is considered between ego-vehicle planning and the predictions of dynamic agents in the environment [2,3].
[1] S. Kim, H. Jeon, J. Choi, and D. Kum, “Improving diversity of multiple trajectory prediction based on map-adaptive lane loss,” arXiv preprint arXiv:2206.08641, 2022.
[2] H. Song, W. Ding, Y. Chen, S. Shen, M. Y. Wang, and Q. Chen, “Pip: Planning-informed trajectory prediction for autonomous driving,” in European Conference on Computer Vision. Springer, 2020, pp. 598–614.
[3] N. Rhinehart, R. McAllister, K. Kitani, and S. Levine, “Precog: Prediction conditioned on goals in visual multi-agent settings,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 2821–2830.
Principales activités
The primary goal of this PhD thesis is to develop an AI-based framework for Trajectory Planning for autonomous driving. This framework leverages data from perception and tracking modules, and interactions with other agents to understand the evolving scene. Its aim is to define safe, optimal trajectories, avoiding obstacles and adhering to traffic rules, without relying on explicit motion models. The focus lies in creating an effective learning strategy to uncover patterns and dependencies in the training data, for both uni-directional and bi-directional planning, addressing challenges such as social interactions among agents, environment topology, and multi-modal trajectory predictions
Compétences
– Engineer with R&D experience or Candidate having a Master in Computer Science,
Robotics or closely related fields.
– Good theoretical and practical background in Robotics and Computer Science
– Good theoretical and practical background in one of the following domains: Robotics, Multi-sensors Perception, Scene Understanding, Parallel computing, Deep Learning and/or Decision-making for safe navigation.
– Good skills in C/C++.
The following qualifications would be an advantage:
– Experience using the Robotics library ROS
– Familiarity with CUDA and Boost libraries, or FPGAs
– Good skills in Linux, system management, python.
– Theoretical knowledge of Bayesian models
– Experience on Deep Learning
– Ability to work as a teammate with other researchers
– Reasonable French and/or English skills (written and spoken)
Avantages
- Subsidized meals
- Partial reimbursement of public transport costs
- Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
- Possibility of teleworking and flexible organization of working hours
- Professional equipment available (videoconferencing, loan of computer equipment, etc.)
- Social, cultural and sports events and activities
- Access to vocational training
- Social security coverage under conditions
Rémunération
-
1and 2nd year: 2082 euros gross salary /month
3rd year: 2190 euros gross salary / month