arrow drop search cross

Appels à candidatures – Contrats postdoctoraux dans le domaine de l'intelligence artificielle et du traitement des données massives


Job announcement for a 18-months postdoctoral position Co-clustering of Massive Longitudinal Data

Project

Longitudinal data that are collected over time are ubiquitous in sociological, behavioral, and medical studies. Longitudinal data are different from functional data and from time series in multiple aspects in- cluding non-regular and infrequent observation intervals, measurement error susceptibility, and presence of missingness due to staggered study entry or dropout. Modern technologies such as smartphones, wearable bands, and smart watches, provide convenient options for collecting such longitudinal data on massive number of individuals as well as massive number of variables over time. In this setting, efficient summaries of information over both dimensions, individuals and variables, are of particular interest to researchers.
This project aims to address the problem of co-clustering – a simultaneous clustering of the in- dividuals and the variables – to summarize complex information contained in such longitudinal data. While co-clustering methods have recently been developed for functional and textual data, extending these methods to longitudinal settings presents particular methodological and practical challenges. Our ongoing work in this direction includes extending co-clustering models to incorporate random effects. The successful candidate will work on further model development, specific for longitudinal data, as well as on derivation and implementation of estimation algorithms, design of numerical experiments and simulation studies, and applications to real data.
Candidate Profile
The ideal candidate will have a strong academic backgrounds in computational and applied statis- tics, and a desire to work on challenging problems in statistical methodology for the social sciences. Experience in model-based clustering is beneficial but not a necessary condition.

Work environment

The candidate will have an office located in the Maasai, INRIA joint-team with Université Côte d’Azur at Sophia-Antipolis, Nice, France. INRIA and UCA campuses offer a vibrant and stimulating work environment. This project is aligned with the objective of the UCA Jedi program, in particular with the Data Science strategic program, and is a funding element of the new UCA/Inria joint team Maasai, which is headed by Charles Bouveyron and which was the result of the UCA Jedi initiative. This project will have strong ties and a possibility of a short visit to the department of Statistics of the University of Washington, Seattle, USA, through collaboration with Elena Erosheva, UCA International Chair in Data Science and Professor of Statistics and Social Work at the University of Washington, Seattle. This project is also part of the Institut 3IA Côte d’Azur that has been recently funded by the French AI initiative.

How to apply

Please email you application to both Charles Bouveyron, , and Elena Erosheva, , by 1st October 2019, including the words “UCA-post-doc” in the e-mail subject line. Women, persons with disabilities, and underrepresented minorities are especially encouraged to apply.
The application should contain:

  1. a CV including a publication list
  2. a letter describing motivation, academic strengths, and related experience to the position
  3. a writing sample (publication or thesis in pdf format)


Job announcement for a 18-months postdoctoral position Modeling and Analysis of Complex and MassiveHeterogeneous Data with Deep Generative Models

August 27, 2019

Project

Artificial intelligence has become a key element in most scientific fields and is now part of everyone life thanks to the digital revolution. Statistical, machine and deep learning methods are involved in most scientific applications where a decision has to be made, such as medical diagnosis, autonomous vehicles or discourse analysis. Such methodologies have also significant implications in fields where the understanding of a phenomenon from data is needed. This is the case for instance in medicine, biology, astrophysics or digital humanities where learning methods allow to recover hidden patterns in the data and to visualize them.

The recent and highly publicized results of artificial intelligence should not hide the remaining and new problems posed by modern data. Indeed, despite the recent improvements due to deep learning, the nature of modern data have brought specific issues. For instance, learning with high-dimensional, atypical (networks, functions, ...), dynamic, or heterogeneous data remains difficult, for theoretical and algorithmic reasons. The recent establishment of deep learning has also open new questions in this context such as: How to learn in an unsupervised or weakly-supervised context with deep architectures? How to learn with evolving and corrupted data?

This project will focus on the development of learning models and algorithms that are able to handle complex heterogeneous data. We in particular target data which involved both structured elements (such as contextual fields, meta-data, ...) and non-structured ones (such as texts, images, ...). For instance, we target the automatic analysis of medical heterogeneous data that can consist in the all available patient data (biological data, MRI/TEP images, functional measures, omic data) and all contextual elements about the patient (clinical path, surgery reports, doctor letters). The analysis of such massive data may result in a significant improvement of both the medical diagnosis and the patient treatment process. Another application field that we profile as a possible application is the analysis of intellectual property data, such as patents or scientific publications. There is indeed a strong need for private companies and public institutes to be able to manage and evaluate the value of their intellectual property. Scientific publications and patents share to be made of both structured elements (contextual fields, meta-data, networks, ...) and non-structured ones (texts, images, schemata, ...).

Although heterogeneous data are indeed parts of the most important and sensitive applications of artificial intelligence, there is a lack of available methods able to deal with such data. Learning methods usually are only able to handle one type of data (continuous data or texts, or images, ...), with eventually some covariates (contextual data). For instance, the most popular method to cluster documents is the latent Dirichlet allocation (LDA) model [3]. For image analysis, convolutional neural networks [4] are nowadays the most efficient algorithms to describe and classify natural images. Beyond single-type data models, proposing unified models for heterogeneous data is an ambitious task, but first attempts (e.g. the Linkage project [1, 2] for instance) on combination of two data types have shown that more general models are feasible. Those models turned out to significantly improve the performances.

In this postdoctoral project, we will address the problem of conciliating structured and non-structured heterogeneous data, as well as data of different levels (individual and contextual data). We ambition in this research project to address two main learning problems: clustering the heterogenous data by taking into account all available information and predicting the value of a subset of key elements (which may be viewed as a regression problem). Interestingly, the two problems may be combined if, for the second situation, we do not have a preselected subset of elements to evaluate. The clustering method that will be proposed may be used to propose such a subset of elements.

Candidate Profile

The ideal candidate will have a strong academic backgrounds in computational, applied statistics and deep learning, and a desire to work on challenging problems in artificial intelligence. Experience in model-based clustering and deep generative models is beneficial but not a necessary condition.

  • Duration: 18 months
  • Expected start date: 1st October 2019.
  • Salary: gross salary per month 3000 EUR (i.e. approx. 2400 EUR after tax) Hosting laboratory: Maasai, INRIA joint-team with Université Côte d’Azur Involved teams:
    • Maasai, INRIA joint-team with Université Côte d’Azur
    • Laboratoire J.A. Dieudonné, UMR CNRS 7351, Université Côte d’Azur Laboratoire I3S, UMR CNRS 7271, Université Côte d’Azur
    • Supervisors:
  • Pr. Charles Bouveyron, Université Côte d’Azur & Chair Inria in Data Science
  • Pr. Frédéric Preciosio, Université Côte d’Azur

Work environment

The candidate will have an office located in the Maasai, INRIA joint-team with Université Côte d’Azur at Sophia-Antipolis, Nice, France. INRIA and UCA campuses offer a vibrant and stimulating work environment. This project is aligned with the objective of the UCA Jedi program, in particular with the Data Science strategic program, and is a funding element of the new UCA/Inria joint team Maasai, which is headed by Charles Bouveyron and which was the result of the UCA Jedi initiative. This project is also part of the Institut 3IA Côte d’Azur that has been recently funded by the French AI initiative.

How to apply

Please email you application to both Charles Bouveyron, , and Frédéric Precioso, , by 1st October 2019, including the words “UCA-post-doc” in the e-mail subject line. Women, persons with disabilities, and underrepresented minorities are especially encouraged to apply.

The application should contain:

  1. a CV including a publication list
  2. a letter describing motivation, academic strengths, and related experience to the position
  3. a writing sample (publication or thesis in pdf format)

 



Deep Neural Networks – Assisted Face Analysis for Health Monitoring

Research field: Computer Vision based Facial Analysis Project-team: STARS, Inria Sophia Antipolis
Funding: Université Côte d'Azur

About Inria and the team

Inria, the French National Institute for computer science and applied mathematics, promotes “scientific excellence for technology transfer and society”. Graduates from the world’s top universities, Inria's 2,700 employees rise to the challenges of digital sciences. With its open, agile model, Inria is able to explore original approaches with its partners in industry and academia and provide an efficient response to the multidisciplinary and application challenges of the digital transformation. Inria is the source of many innovations that add value and create jobs.

Team
The STARS research team combines advanced theory with cutting edge practice focusing on cognitive vision systems.

Team web site
https://team.inria.fr/stars/

Mission

The Post Doc position is within the framework of the UCA Postdoctoral Fellowship DNN4HM.

  • Starts fall 2019
  • Deadline for applications: November 2019. Nevertheless, the application may be closed before the deadline, if a satisfying candidate is found

Job description

The Inria STARS team is seeking for a Post Doctoral researcher with strong background in computer vision, deep learning, machine learning and applied mathematics.
The candidate is expected to conduct research toward finding fundamental principles in computer vision.

The proposed research project Deep Neural Networks for Health Monitoring (DNN4HM) aims to provide computer vision methods for facial expression recognition in patients with Alzheimer's disease (AD). Most importantly, the work seeks to be a part of a paradigm shift in current healthcare in finding effective, cost-efficient and objective measures to assess different therapy treatments, as well as to enable automated human-computer interaction in remote large-scale healthcare frameworks. First objective is to propose a model for facial expression recognition in healthcare settings, where we aim to develop a general approach, where expression models are averaged over the set of subjects and a subject based approach, where we consider subjects individually and classifiers are trained using general expression analysis. Second objective is to fuse and integrate different methods with respect to different user needs and environmental requirements including hospital configuration and sensor setting, as well as ethical issues.
To reach our objectives, we will be designing new models of machine learning and computer vision, specifically deep neural networks and methods for analysis of videos comprising human faces, investigating new attention and partitioning mechanisms, and exploring alternatives to supervised learning. Such methods will analyze appearance and dynamics of face and body, towards recognition of identity, gender, age, emotions, behavior, as well as mental and social states of humans in unconstrained settings. Such dynamics, which include facial expressions, visual focus of attention, hand and body movement, will provide cues to a new class of tools that will be instrumental for efficient analysis of elderly subjects for automated healthcare.

Skills and profile

Candidates must hold a Ph.D. in Computer Science or a closely related discipline. Candidates must also show evidence of research productivity (e.g. papers, patents, presentations, etc.) at the highest level.
The candidate must be grounded in the basics of computer vision, have solid mathematical and programming skills.
The candidate must be committed to scientific research and strong publications.

Advantages

  • Inria Sophia Antipolis is ideally located in the heart of the French Riviera, inside the multi-cultural silicon valley of Europe
  • Strong social / medical benefits
  • Restaurant on site
  • Financial participation for public transport
  • Social and sporting activities
  • French courses

Additional Information

  • Duration: 18 months
  • Targeted hiring date: Fall 2019
  • Location: Inria Sophia Antipolis, France

Application

To apply, please email the following documents to Antitza Dantcheva (), indicating “UCA – Post Doc” in the e-mail subject line:

  • Cover letter
  • CV
  • List of publications
  • Future plan of research with possible links to the project
  • Contact information for at least two references who can provide recommendation letters upon request.

The submission deadline is November 2019. Nevertheless, the application may be closed before the limit date, if a satisfying candidate is found.

Please do not hesitate to contact us for any inquiry.

Inria's disabilities policy: All positions at the institute are open to disabled people.

Security and defense procedure
In the interests of protecting its scientific and technological assets, Inria is a restricted-access establishment. Consequently, it follows special regulations for welcoming any person who wishes to work with the institute. The final acceptance of each candidate thus depends on applying this security and defense procedure.



Extraction of curvilinear structure networks in image data using an innovative deep learning approach: application to fracture and fault network extraction from satellite data

Application before July 31, 2019

18 months post-doc position on project “Extraction of curvilinear structure networks in image data using an innovative deep learning approach: application to fracture and fault network extraction from satellite data”

Project

Curvilinear structure networks are widespread in both nature and anthropogenic systems, ranging from angiography, earth and environment sciences, to biology and anthropogenic activities. Recovering the existence and architecture of these curvilinear networks is an essential and fundamental task in all the related domains. At present, there has been an explosion of image data documenting these curvilinear structure networks. Therefore, it is of upmost importance to develop numerical approaches that may assist us efficiently to automatically extract curvilinear networks from image data.

In recent years, a bulk of works have been proposed to extract curvilinear networks. However, automated and high-quality curvilinear network extraction is still a challenging task nowadays. This is mainly due to the network shape complexity, low-contrast in images, and high annotation cost for training data. To address the problems aroused by these difficulties, this project intends to develop a novel, minimally-supervised curvilinear network extraction method by combining deep neural networks with active learning, where the deep neural networks are employed to automatically learn hierarchical and data-driven features of curvilinear networks, and the active learning is exploited to achieve high-quality extraction using as few annotations as possible. Furthermore, composite and hierarchical heuristic rules will be designed to constrain the geometry of curvilinear structures and guide the curvilinear graph growing.

The proposed approach will be tested and validated on extraction of tectonic fractures and faults from a dense collection of satellite and aerial data and “ground truth” available at the Géoazur laboratory in the framework of the Faults_R_Gems project co-funded by the University Côte d’Azur (UCA) and the French National Research Agency (ANR). Then we intend to apply the new automatic extraction approaches to other scenarios, as road extraction in remote sensing images of the Nice region, and blood vessel extraction in available medical image databases.

Candidate profile

Strong academic backgrounds in Stochastic Modeling, Deep Learning, Computer Vision, Remote Sensing and Parallel Programming. A decent knowledge of Earth and telluric features (especially faults) will be appreciated.

At UCA, Géoazur and Inria we seek to increase the number of women in areas where they are under-represented and therefore we explicitly encourage women to apply. We are also committed to increasing the number of individuals with disabilities in our workforce and therefore we encourage applications from such qualified individuals.

Post-doc salary and conditions

Duration: 18 months

Starting date: between September 1st and November 1st, 2019
Salary: gross salary per month 3000 EUR (i.e. approx. 2400 EUR net)
Hosting laboratory: GEOAZUR, Sophia Antipolis (https://geoazur.oca.eu/fr/acc-geoazur) Advisors: Drs. Isabelle MANIGHETTI (Géoazur) and Josiane ZERUBIA (Inria-SAM). See: https://www.oca.eu/fr/isabelle-manighetti http://www-sop.inria.fr/members/Josiane.Zerubia/index-eng.html

Work conditions: Tight collaboration between Géoazur and Inria-SAM (http://www- sop.inria.fr). Position located at Géoazur, with research discussions planned twice a week at Inria Sophia.

How to apply

Dead-line to apply: July 31, 2019

Please email a full application to both Isabelle Manighetti (manighetti@geoazur.unice.fr) and Josiane Zerubia (josiane.zerubia@inria.fr), indicating “UCA-AI-post-doc” in the e-mail subject line. The application should contain:
- a motivation letter demonstrating motivation, academic strengths, and related experience to the position
- CV including publication list
- at least two major publications in pdf - minimum 2 reference letters

Contacts :

Isabelle Manighetti :
Josiane Zerubia :



Targeting ion transport in Cancer Metabolism and Invasion

Our laboratory, LP2M UMR7370 is recruiting an international post doctoral fellow for a period of 18 month starting if possible as early as September 2019.

The candidate will work on a transdisciplinary project aimed at deciphering the relations between the activities of Na+/H+ exchangers, Na+ Channels and Na/K ATPase in the field of cellular metabolic reprogrammation, motility and invasiveness. The experimental approaches will combine ion transport and cellular measurements with transcriptomics and metabolomics and will therefore involve the analysis of large sets of data.

The candidate will benefit from the infrastructures of LP2M and of the surrounding technical platforms: Electrophysiology, atomic absorption spectrometry analytical HPLC, flow cytometry, videomicroscopy, genomics, imaging, metabolomics... The candidate will also work in collaboration with colleagues from the Maths/Physics and computing departments for data analysis and model construction. 

The candidate will have to possess a high-level expertise in ion transport, cellular biology and metabolism to perform the experimental parts of the project. An expertise in mathematical and or computational approaches of biology is not mandatory but will be appreciated.

Additional information:

Approximate net salary : 2400€/month

Laboratory website: http://unice.fr/lp2m/fr

Please send a detailed CV and two names of scientific colleagues or former supervisors, for reference to Pr. Laurent Counillon : Laurent.Counillon@univ-cotedazur.fr



Analysis and biological interpretation of multi-omics data in the Human Lung Atlas

Research field: Single cell genomics, cell biology, personalized medicine, machine learning

Team: Pascal Barbry’s laboratory, IPMC Sophia Antipolis (CNRS & Université Côte d’Azur ), France
Funding: Université Côte d'Azur

About IPMC and the team

Founded in 1989 in the Scientific Park of Sophia Antipolis (French Riviera), IPMC is a joint laboratory between the  Centre National de la Recherche Scientifique (CNRS) and Université Côte d’Azur. It explores with industrial and academic partners original approaches and models at the front-end of research in biology, genomics, molecular and cellular pharmacology.

Team
The « Physiological Genomics of the Eukaryotes » research team combines state-of-the-art expertise in genomics with cutting edge practice focusing on the airway epithelium. The laboratory is fully equipped in sequencers, robots, microscopes and FACS for all wetlab developments, with access  to ad hoc informatic resources for all in silico treatments.

Team web site
https://www.ipmc.cnrs.fr?page=barbry

Mission

The Post Doc position is within the framework of the UCA Postdoctoral Fellowship.

Job description

The group of Pascal Barbry is hiring a Post-Doctoral researcher with strong background in bioinformatics, machine learning and applied mathematics. The candidate will work on the integration of data from several single-cell experiments and techniques already or currently deployed by the host laboratory, with the final goal of unraveling the sequences of molecular events controlling the balance between distinct airway cell types during health and disease. By deciphering the epithelial-related molecular mechanisms it is anticipated that more efficient therapies will be possibly set up.

Our team contributes to the international Human Lung Atlas seed network, which works on the construction of the first comprehensive human lung cell atlas. The Post-Doctoral researcher will analyze the datasets produced in our laboratory and by collaborators in the framework of this consortium, including datasets currently produced by our laboratory and studying alterations of the airways in severe asthma (remodeled airway epithelium, hyperplasia of mucus-secreting cells, etc). The postdoc will analyze with dedicated computational approaches a huge database incorporating a full genome investigation on more than 100 000 cells.

The main challenge of this multiscale project will be to establish genome-wide RNA expression, splicing and genome accessibility through short and long reads single cell sequencing. The resulting datasets will be also merged with spatial transcriptomics data. One idea will be to establish in a multidimensional gene expression space possible trajectories between a specific progenitor cell and different types of terminally differentiated cells. Exact spatial distribution of cell types in the biological tissue will then be connected to this reconstructed gene expression space.

The post-doctoral candidate will work on four subtasks:

Skills and profile

Advantages

Additional Information

Application

To apply, please email the following documents to Pascal Barbry (barbry@ipmc.cnrs.fr), indicating “UCA – Post Doc” in the e-mail subject line:

The submission deadline is November 2019. Nevertheless, the application may be closed before the limit date, if a satisfying candidate is found.

Please do not hesitate to contact us for any inquiry.

IPMC's disabilities policy: All positions at the institute are open to disabled people.

Security and defense procedure: In the interests of protecting its scientific and technological assets, IPMC is a restricted-access establishment. Consequently, it follows special regulations for welcoming any person who wishes to work with the institute. The final acceptance of each candidate thus depends on applying this security and defense procedure.

Recent publications from the laboratory:

  1. Sandra Ruiz García, Marie Deprez, Kevin Lebrigand, Agnès Paquet, Amélie Cavard, Marie-Jeanne Arguel, Virginie Magnone, Marin Truchi, Ignacio Caballero, Sylvie Leroy, Charles-Hugo Marquette, Brice Marcet, Pascal Barbry, Laure-Emmanuelle Zaragosi. 2018. Novel dynamics of human mucociliary differentiation revealed by single-cell RNA sequencing of nasal epithelial cultures. Biorxiv https://www.biorxiv.org/content/10.1101/451807v1
  2. Diego R. Revinski, Laure-Emmanuelle Zaragosi, Camille Boutin, Sandra Ruiz-Garcia, Marie Deprez, Olivier Rosnet, Virginie Thomé, Olivier Mercey, Agnès Paquet, Nicolas Pons, Brice Marcet, Laurent Kodjabachian, Pascal Barbry. 2018. CDC20B is required for deuterosome-mediated centriole production in multiciliated cells. Nature Communications. 9(1):4668
  3. Lisa Giovannini-Chami, Agnès Paquet, Céline Sanfiorenzo, Nicolas Pons, Julie Cazareth, Virginie Magnone, Kévin Le Brigand, Benoit Chevalier, Ambre Vallauri, Valérie Julia, Sylvie Leroy, Brice Marcet, Charles-Hugo Marquette, Pascal Barbry. 2018. The “one airway, one disease” concept in light of Th2 inflammation. Eur Respir J. 52(4). pii: 1800437.
  4. Olivier Mercey, Alexandra Popa, Amélie Cavard, Agnès Paquet, Benoît Chevalier, Nicolas Pons, Virginie Magnone, Joséphine Zangari, Patrick Brest, Laure-Emmanuelle Zaragosi, Gilles Ponzio, Kevin Lebrigand, Pascal Barbry, Brice Marcet. Characterizing isomiR variants in the microRNA-34/449 family. 2017. FEBS Letters. 591(5):693-705.
  5. Marie-Jeanne Arguel, Kevin LeBrigand, Agnès Paquet, Sandra Ruiz-Garcia, Pascal Barbry, Rainer Waldmann. A cost effective single cell RNA sequencing approach for the Fluidigm C1. 2016. Nucleic Acids Research. gkw1242
  6. Alexandra Popa, Kevin Lebrigand, Nicolas Nottet, Agnès Paquet, Karine Robbe-Sermesant,  Rainer Waldmann, Pascal Barbry. RiboProfiling: a Bioconductor package for Ribo-seq data processing. F1000.
  7.   Alexandra Popa, Kevin Lebrigand, Pascal Barbry , Rainer Waldmann. Pateamine A-sensitive ribosome profiling reveals the scope of translation in mouse embryonic stem cells. BMC Genomics. 17(1):52
  8.   Chevalier B, Adamiok A, Mercey O, Revinski DR, Zaragosi LE, Pasini A, Kodjabachian L, Barbry P, Marcet B. "miR-34/449 control apical actin network formation during multiciliogenesis through small GTPase pathways". 2015. Nature Communications. 6:8386.
  9. Lisa Giovannini-Chami, Brice Marcet, Chimène Moreilhon, Benoît Chevalier, Marius I. Illie, Kévin LeBrigand, Karine Robbe-Sermesant, Thierry Bourrier, Jean-François Michiels, Bernard Mari, Dominique Crénesse, Paul Hofman, Jacques de Blic, Laurent Castillo, Marc Albertini, Pascal Barbry. Distinct epithelial gene expression phenotypes in childhood respiratory allergy. European Respiratory Journal. 39(5):1197-205.
  10. Brice Marcet, Christelle Coraux*, Benoît Chevalier*, Guillaume Luxardi*, Laure-Emmanuelle Zaragosi, Karine Robbe-Sermesant, Thomas Jolly, Bruno Cardinaud, Chimène Moreilhon, Lisa Giovannini-Chami, Philippe Birembaut, Rainer Waldmann, Laurent Kodjabachian, Pascal Barbry. miR-449 microRNAs trigger vertebrate multiciliogenesis through direct repression of the Notch ligand Delta-like 1. Nature Cell Biology. 13(6):694-701