36 Postdoctoral Fellowship at Inria, France

The postdoctoral position addresses the challenge to build a “human-centered digital word” with innovative interactive systems providing “adaptive and personalized interactions with humans” for supporting “their wellbeing and health”. Additionally, this PhD program is primarily based on interdisciplinary approach of individuals’ intrinsic motivation (curiosity) as critical ingredient for synergistic human-computer interactions. Today, very few studies provide information on the impact of intrinsic motivation and curiosity on the inter-individual variability of learning across life span or on their neuro-protective role against the aging effects or attentional disorders.

With the help of the PASTA team (Inria) and the Archimède team (Université Strasbourg & Université de Haute-Aslace), the recruited person will be taken to test several techniques of factor analysis and text classification in collaboration with historian to create a lexicon of ideas, with a view toward automatization through reinforcement learning techniques.

The goal of this post-doc is to contribute to the state of the art of the research in reinforcement learning. The work will be essentially at the fundamental level: design of new algorithms, analysis of the behavior of algorithms, implementation and experimental study and assesment, …

The present postdoctoral project is part of a collaborative project between the Atlantis project-team from the Inria Research Center at Université Côte d’Azur,  (2) Thales R&T in Palaiseau, France, which is the main multidisciplinary research center of Thales group, and one of the main world players in professional electronics (3) and Nanoe in Ballainvilliers, France, whcih is  a SME (Small and Medium Entreprise) specialized  in development and fabrication of innovative ceramic materials for the industries.

Atlantis is  a joint project-team  between Inria and  the Jean-Alexandre Dieudonné Mathematics Laboratory at  Université Côte d’Azur. The team  gathers applied mathematicians and  computational scientists who are collaboratively undertaking  research activities aiming at the design, analysis, development and  application of innovative numerical methods for systems of  partial differential equations (PDEs) modelling nanoscale light-matter interaction problems. In this context, the team is  developing  the   DIOGENeS  [https://diogenes.inria.fr/]  software suite,  which  implements  several Discontinuous  Galerkin  (DG)  type methods tailored to the systems  of time- and frequency-domain Maxwell equations  possibly coupled  to  differential  equations modeling  the behaviour of propagation  media at optical frequencies.  DIOGENeS is a unique  numerical   framework  leveraging   the  capabilities   of  DG techniques  for  the simulation  of  multiscale  problems relevant  to nanophotonics and nanoplasmonics.

The candidate will be hired within the SITAR project (Impact of Information Structures on Service Pricing), funded by the call for project STaRS of the Hauts-de-France région. The candidate will be integrated within the INOCS team, at Inria Lille-Nord Europe. The offer is defined on a 12 months basis, but can be extended up to 18 months. The candidate will be encouraged to take part in international collaborations with the University of Athens and/or with Polytechnic Montréal. The goal of this post-doc is to propose new methods and algorithms to price services targeted at end users, in context of competition where the information available might be limited due to privacy constraints imposed by the end users, or strategic.

IRSN is public expert, with industrial and commercial activities, in nuclear and radiation risks, and its activities cover all the related scientific and technical issues.The Institute is supervised jointly by the French Minister of the Ecological transition, the French Minister of Defense, and the French Ministers of Energy transition, Research and Health.
The Radiobiology of Accidental Exposure Laboratory (LRAcc) develops operational methods and tools used for the diagnosis and prognosis of radiation-induced organ damage. Among other things, it proposes dose assessment approaches based on radiation-induced damage based on radiation-induced damage in the a posteriori reconstitution of absorbed doses (biological dosimetry).

The overall goal of this project is to investigate how computer tools can help capturing and reproducing the typicality of traditional 2D animations. More precisely, we plan to address the following two challenges:

1. How to interactively generate in-betweening frames from a sparse sequence of rough drawings to swiftly explore different animation choices and motion designs.
2. How to stylize the geometry and motion of a 3D animation to allow its seamless integration into a 2D animation.

These two objectives target tasks that are currently tedious to perform by artists; they currently rely on manual techniques or ad‐hoc solutions that severely limit artistic creativity and drastically increase production times. The solutions that we will develop in the project should be compatible with regular animation workflows and should provide adequate controls to artists that should remain at the center of the computer‐human interaction loop.

The Inria STARS team is seeking for a Post Doctoral researcher with strong background in computer vision. We are especially interested in candidates with background in computer vision, deep learning, machine learning and applied mathematics. The candidate is expected to conduct research related to generative models, and specifically generation of facial images and videos, i.e., deepfakes. The recruitee will mainly work with Antitza Dantcheva, but beyond that, he/she will be able to collaborate with members of the STARS team. Support for traveling e.g. to conferences or other scientific meetings is provided.

3D printing using fused filament deposition (FDM) [1] has many fields of application including prototyping, maintenance and custom manufacturing. Recently, during COVID, large-scale emergency production of innovative solutions for crisis situations has been made possible by this technology, which is easily deployable and allows on-site production. Within the Ofast3D exploratory project, we aim to increase the production capacity of players using 3D printing by fused filament deposition without modifying their infrastructures. This project is a collaboration between the PACAP team for the compilation and optimization aspects for performance, MFX for its experience in 3D printing and MimeTIC for the planning and scheduling aspects.

The mutual low-latency objective for both Data Stream Processing (DSP) and Fog environments has resulted in a continuous growth of DSP deployments on Fog environments. The success of DSP deployments in the Fog relies on operators placements and the ability to sustain low latency. Accordingly, much work have focused on placement strategies across Edge-servers or across hybrid Cloud and Edge environments. Previous efforts have focused on reducing the volume of communication overhead between nodes (inter-node communication) and dividing the computation between edge servers and clouds. Unfortunately, they are oblivious to (1) the dynamic nature of data streams (i.e., data volatility and bursts) and to (2) the bandwidth and resource heterogeneity in the Edge, which negatively affects the performance of stream data applications.

L’objectif est le développement de méthodes symboliques et numériques efficaces pour calculer les périodes de variétés algébriques.

Nous cherchons des canditats intéressés par la modélsisation probabiliste des grands réseaux de communications.

Radiotherapy (RT) is one of the most important treatments of primary brain tumors, of which 60% are high grade. However, its potential neurotoxicity on the central nervous system is a highly relevant clinical issue. It is also part of the priority research questions in radiation protection, regarding the identification and the prevention of non-cancer side effects related to the use of ionizing radiation (IR) for therapeutic purposes. Currently, the most frequent and threatening mid to long-term neurotoxic complication of brain RT is cognitive dysfunction related to radiation-induced leukoencephalopathy (RIL). Image-based biomarkers for RIL include diffuse supratentorial white-matter lesions (WML), ventricular dilatation, and brain atrophy (BA). The associated cognitive impairments can dramatically reduce the quality of life for long-term survivors. The neurocognitive status is also an important end-point in clinical trials. However, the underlying physiopathology of radiation-induced neurotoxicity in normal tissues and organs is poorly understood as well as its potential links with the initiation and temporal progression of specific cognitive dysfunctions.

The goal is to study the best parallel and quantum algorithms for lattice problems, as well as design new tools to select safe parameters for lattice-based cryptography.

 

(18) Postdoctoral Fellowship Position

Postdoc Fellowship Position summary/title: Post-Doctoral Research Visit F/M Postdoc in isogeny based cryptography

Deadline :  2022-11-30

View details & Apply

 

Ce poste est proposé dans le cadre d’un projet de recherche dont l’objectif est de créer un entrepôt de données pour intégrer et analyser les données des partenaires du projet liées à l’antibiorésistance (résistance aux antibiotiques). Le développement de l’entrepôt de données est en cours de réalisation avec un ingénieur de l’équipe Zénith de l’Institut national de recherche en sciences et technologies du numérique (Inria). L’équipe Zenith, experte en gestion des données scientifiques, a une solide expérience dans le traitement et analyse de grandes quantités de données produites par différentes méthodes. Après la mise en place de l’entrepôt, nous aurons besoin d’analyser les données pour extraire de nouvelles connaissances faisant le lien entre la santé humain, la santé animale, et l’environnement. Par exemple la relation entre la consommation des antibiotiques et la résistance des bactéries dans le temps, ou la similarité de résistance entre l’homme et l’animal et l’impact des conditions environnementales. Parmi les pistes, nous envisageons d’utiliser des techniques d’analyse de séries temporelles et aussi des méthodes d’analyse de données massives.

In this project we consider more particularly the context of the use case of CiGri [1], a lightweight, scalable and fault-tolerant grid system that exploits the unused resources of a set of computing clusters. Harvesting idle resources is done by submitting jobs from Bag-of-Tasks applications as best-effort, with the lowest priority to the clusters in order to maximize their use. Such best-effort jobs can be killed when premium computations need resources. However, the allocation of resources to best-effort jobs introduces disturbances for the premium users of the cluster: increased startup delays due to the harvesting of resources, longer scheduling decisions delays, degraded performances of the distributed file system, … Therefore, the submission must be regulated in a feedback loop with a Control Theory approach [4] in order to minimize these perturbations.

The AI Verse technology (http://www.ai-verse.com/) is devised to create infinitely random and semantically consistent 3D scenes. This creation is fast, consuming less than 4 seconds per labeled image. From these 3D scenes, the system is able to build quality synthetic images that come with rich labels that are unbiased unlike manually annotated labels.
As for real data, no metric exists to evaluate the performance of the synthetic datasets to train a neural network. We thus tend to favor the photorealism of the images but such a criterion is far from being the best. The current technology provides a means to control a rich list of additional parameters (quality of lighting, trajectory and intrinsic parameters of the virtual camera, selection and placement of assets, degrees of occlusion of the objects,  choice of materials, etc). Since the generation engine can modify all of these parameters at will to generate many samples, we will explore optimization methods for improving the sampling quality.

The postdoctoral fellow will develop these ideas in the software projects developed by the Inria TADaaM team. This notably includes the hwloc library for low-level hardware management and operating system interaction. Higher level ideas will be designed in collaboration with the RWTH Aachen team, with the goal of proposing ideas to the OpenMP standard committee. Modifications of the LLVM compile and runtimes might be used as a way to identify sensitivity and apply placement heuristics.

This Post-doc subject is proposed in the context of the ROAD-AI Challenge involving the Cerema and the Inria Team ACENTAURI. ACENTAURI is a robotic team located in Sophia Antipolis that studies and develop intelligent, autonomous and mobile robots that collaborate between them to achieve challenging tasks in dynamic environments. The team tackle perception, decision and control problems for multi-robot collaboration by proposing an original hybrid model-driven/data driven approach to artificial intelligence and by studying efficient algorithms. The team focus on robotic applications like environment monitoring and transportation of people and goods. In these applications, several robots will share multi-sensor information eventually coming from infrastructure. The effectiveness of the proposed approaches are demonstrated on real robotic systems like cars AGVs and UAVs together with industrial partners.

ACENTAURI est une équipe de robotique dirigée par Ezio MALIS qui étudie et développe des robots autonomes et intelligents qui collaborent entre eux pour réaliser des tâches difficiles dans des environnements complexes et dynamiques. L’équipe aborde les problèmes de perception, de décision et de contrôle pour la collaboration multi-robot en proposant une approche hybride originale de l’intelligence artificielle basée sur les modèles et les données et en étudiant des algorithmes efficaces. L’équipe se concentre sur les applications telles que les systèmes multi-robot rondiers pour la surveillance de l’environnement et le transport de personnes et de marchandises. Dans ces applications, plusieurs robots partagent des informations multi-capteurs provenant éventuellement de l’infrastructure. L’efficacité des approches proposées est démontrée sur des systèmes robotiques réels comme les voitures et les drones en collaboration avec des partenaires industriels.

The goal of this 24-month postdoc is to automate monitoring, with sensor-based intelligence to validate the factory activities. Fixed or mobile sensor systems are planned (cameras and laser scanners). The first problem to be studied is the placement of the sensors which must cover the whole area to be monitored. The second problem is that there are differences between the parts designed in CAD and those assembled. In addition, the “as-designed” model may overlap with the “as-built” model, and equipment may appear late. The equipment of the construction site can mask the perception of reality. The required accuracy must be sufficient to validate the presence and valid mounting of equipment. The detection of assembly errors will be performed by comparing the geometric model with the reality reconstructed in 3D from Lidar point clouds or by photogrammetry, or by using images from the camera. Machine learning approaches will be used to correlate the acquired images with the digital mockup projected onto the images. The full-scale experiments will take place on Naval Group’s construction sites in Lorient. Travel will be required for system development, data acquisition and demonstrations.

The selected applicant will work within the WIDE research team at Inria of the University of Rennes (https://team.inria.fr/wide/). WIDE focuses on large-scale distributed systems and applications, as well as on privacy, and distributed machine learning. This postdoctoral position is funded by the SOTERIA H2020 project (https://www.soteria-h2020.eu/) which aims to develop a user-centric, citizen-driven tool to help European citizens to manage their personal data. The project therefore involves close collaboration with French international partners such as AriadNext by IDNow, Idiap Switzerland, COSIC Team @ KU Leuven, a complete list is available at https://www.soteria-h2020.eu/partners/. 

In this PostDoc, we focus on a class of metamaterials, called auxetics ‎[1]. Auxetic materials have a counter-intuitive deformation behavior: instead of becoming thinner when stretched, they become thicker. Auxetic materials are known to exhibit properties important in applications as diverse as biomedical engineering, aerospace industry, or protective sports gear, among others. We are researchers from three Computer Graphics team, that have developed complementary approaches for generating 2D planar auxetic networks, either by pruning a dense network ‎[6], or by conducting a parametric growth process ‎[3]‎[4][5]. This PostDoc aims to extend our previous results for constructing 3D structures and porous surfaces with tailored mechanical behavior, with an emphasis on auxeticity. Optimization of shape parameters and numerical simulations are part of the project in order to test and optimize the elastic response of the developed geometric structures. An existing numerical simulation code developed in ‎[6] will have to be extended to 3D. Great care will be taken to control the mechanical properties’ directional characteristics rigorously. Once the new structure (metamaterial) is computed, it can be used as a mesoscale structure and fill a 3D geometric object with it.

This position is funded by Inria following a specific call “Actions Exploratoires” (fr.: “Exploratory Actions”) within the project DATA4US and will be hosted in the Privatics team at Sophia Antipolis. The Privatics team is a research group affiliated to Inria based in Grenoble, Lyon and Sophia Antipolis. Privatics follows a multidisciplinary approach in considering the scientific and technical issues, but also economic, legal and social aspects of privacy. The team has expertise in the identification of privacy issues, anonymization techniques and sanitization database and design of Privacy Enhancing Technologies (PETs). Privatics has a long history of contributing to Standards Developing Organisations (IETF and IEEE), to the public consultations of the European Data Protection Board and EU authorities and has a strategic partnership with the French Data Protection Authority (CNIL).

Some of the musical wind instruments stored in museums cannot be played for conservation reasons, and are sometimes incomplete or damaged (lost or broken parts, …). It is however possible to perform non destructive images (geometric, acoustical measurements, tomography, etc) giving access to the inner shape of the air column. From this data, modeling techniques can compute the physical, acoustical and vibratory functionning of museum instruments allowing to guess musical properties (such as the tuning) or the context of use. In order to add to these characteristics a numerical simulation of the sound that could be produced by such an instrument, it is necessary to estimate the values of the parameters of the equation modelling the player (supply pressure, etc.), which are suitable for playing the instrument considered. As a first step towards this goal, this project aims to find the set of parameters allowing the reproduction of a real sound with a physics-based sound synthesis. This study will focus on the recorded sound of a replica trumpet from the Besson collection of the Musée de la Musique – Philharmonie de Paris.

This post-doctoral position will be supported by the TIP project on Transparent artificial Intellgence preserving Privacy. This project will be jointly funded by I-Site, INRIA, U-Lille and MEL. This is a project in the MAGNET team (INRIA-Lille, https://team.inria.fr/magnet/). While this position will be in the MAGNET team in Lille, we will collaborate with users (e.g. medical research groups in CHU-Lille) for the validation and exploitation of the work.

This project is about the detection of privacy issues in mobile system. Depending on the skills and background of the candidate, the work will focus on mobile application (Android) or on traces of wireless traffic (Wi-Fi, Bluetooth, …). The objective is to automatically identify undiscovered privacy issues based on a training set of known problematic cases. Indeed, privacy issues are regularly found in mobile applications [1] and in wireless network implementations [2], and tools to efficiently and quickly identify them are required.

The postdoc, of the duration of 24 months, will take place in the framework of the Défi Inria EQIP (Engineering for quantum information processors). It will be supervised by two Inria teams CAGE (Inria center of Paris), McTAO (Inria center of Sophia-Antipolis). The postodc is expected to spend one year in Paris working as a member of the team CAGE and one in Sophia-Antipolis as a member of the team McTAO. Frequent travels of the postdoc between the two teams are expected (and will be convered by the two teams).