The LISN (Interdisciplinary Laboratory of Digital Sciences) of Paris-Saclay University and the ILLS laboratory (International Laboratory on Learning Systems) of McGill University jointly with ETS Montreal and MILA (Quebec Institute of Artificial Intelligence) in Canada(More information: https://www.mcgill.ca/channels/fr/channels/news/un-laboratoire-international-dintelligence-artificielle-voit-le-jour-montreal-339253 ) are looking for a candidate to carry out a joint supervision thesis.
Title of the thesis: Detection of anomalies in texts by statistical analysis of lexical distributions (Statistical analyzes of lexical distributions with an application to anomaly detection in natural texts)
The thesis will be supervised by François Yvon (LISN, CNRS), Pablo Piantanida (ILLS, CNRS - Paris-Saclay University).
To consult the complete subject and submit an application: https://emploi.cnrs.fr/Offres/Doctorant/UMR9015-FRAYVO-009/Default.aspx
The application file must contain: • Detailed CV, • Cover letter, • Details of transcripts (including M1 and M2), • Elements of bibliography or personal achievements relating to a research activity (eg master's project, subject of research internship, etc.), • 2 letters of recommendation.
Application deadline: July 14, 2022
Please distribute this attached document to your students and colleagues likely to distribute in turn.
We are opening an IRD IE position in machine learning applied to physical AND biogeochemical oceanography at LOPS
Deadline July 20
See : https://lnkd.in/ecfaig-w and especially : https://lnkd.in/eutQDVeZ
https://external-careers.jobs.unsw.edu.au/cw/en/job/509128/research-associate-in-cognitive-robotics-and-knowledge-representation
Job no: 509128 Work type: Full time Location: Sydney, NSW Categories: Post doctoral research fellow
More Information: https://external-careers.jobs.unsw.edu.au/cw/en/job/509128/research-associate-in-cognitive-robotics-and-knowledge-representation
The Opportunity As a Research Associate you will conduct research in the field of Cognitive Robotics and Knowledge Representation & Reasoning. This position will give you the opportunity to develop your research skills and outputs, contributing to the writing of scientific papers and reports for international conference and journals, participating in conferences and workshops, and actively engaging with industry partners.
The role of Research Associate reports to Professor Maurice Pagnucco, Deputy Dean (Education) and has no direct reports.
• Level A, Salary - $99,325 to 106,237 per annum + 17% superannuation • Full time • Fixed-term contract – 3 years • Location: Kensington – Sydney, Australia
About UNSW UNSW is currently implementing a ten-year strategy to 2025 and our ambition for the next decade is nothing less than to establish UNSW as Australia’s global university. This position is within the School of Computer Science and Engineering.
Skills & Experience: • PhD (or soon to be awarded) in knowledge representation and reasoning (artificial intelligence) or related area. • Demonstrated ability to conduct independent research with limited supervision. • Demonstrated track record of publications and conference presentations relative to opportunity.
• Demonstrated ability to work in a team, collaborate across disciplines and build effective relationships. • Strong interpersonal skills with demonstrated ability to communicate and interact with a diverse range of stakeholders and students. • An understanding of and commitment to UNSW’s aims, objectives and values in action, together with relevant policies and guidelines. • Ability and capacity to implement required UNSW health and safety policies and procedures.
Additional details about the specific responsibilities for this position can be found in the position description.
To Apply: Please click the apply now button and submit your CV, Cover Letter and Responses to the Skills and Experience. Please note applications will not be accepted if sent to the contact listed below.
Contact: Professor Maurice Pagnucco - Deputy Dean (Education) Email: [email protected]
Eugene Aves – Talent Acquisition Consultant Email: [email protected]
Applications close: 11:55 pm Sydney time on Wednesday, 29th June 2022
UNSW aspires to be the exemplar Australian university and employer of choice for people from diverse backgrounds.
UNSW aims to ensure equality in recruitment, development, retention and promotion of staff, and that no-one is disadvantaged on the basis of their gender, cultural background, disability, sexual orientation or identity. We encourage everyone who meets the selection criteria to apply.
Position Description: https://secure.dc2.pageuppeople.com/apply/841/gateway/default.aspx?sData=E7tTKDEyyWPyLsAhbioSII1xXoCpeOozuBBdaSkT-laczr_Amm6xKZk1R-bRPAwQLQbeK55OFDkrIEVP1QmLTfWkhN9IM9zo273NvMK7SnnJWaAZDxr4qjcm7yzGsEl08T8AQzWhPzwUJ_Y_7slSm6gJbwegdiERcvcYtmfZmqo~
Advertised: 02 Jun 2022 AUS Eastern Standard Time
Applications close: 29 Jun 2022 AUS Eastern Standard Time
The Pl@ntNet team is looking for a data scientist with strong skills in python and pytorch development.
You will work on exciting AI applications for biodiversity monitoring as a research engineer at Inria (in Montpellier antenna).
More information here: https://jobs.inria.fr/public/classic/en/offres/2022-05126
This thesis is part of the Digital Transformation of Maritime Transport (TNTM) project funded by the PIA (Investment Program for the Future). This project is supported by a consortium made up of industrial and academic partners, including CMA CGM, a world leader in the maritime transport sector. The objective is to optimize logistics flows with the aim of competitiveness and environmental preservation, in a context where data evolves in real time.
Maritime transport has been largely globalized in recent years. A model of mega alliances has also been set up to pool fleets of ships as well as port coverage on several continents. The main players operate between 500 and 700 ships with a total capacity of between 3 and 4 million TEUs (Twenty Foot Equivalent, the unit of measurement for containers), serve around 520 commercial ports around the world and manage more than 200 lines. maritime. They can call on additional third-party "feeders" to serve certain ports in order to top up volumes on the main lines. Furthermore, maritime transport is an activity subject to very restrictive national (depending on the country of call) and international (imposed by the International Maritime Organization (IMO)) regulations. Thus the regulations on CO2 emissions are constantly being strengthened, with the aim of reducing emissions from ships in operation. In addition, the economic context is constantly and rapidly changing, in particular due to the years of crisis linked to COVID, which were followed by a very strong recovery in activity.
In general, ships carry out rotations during which containers are loaded and unloaded at the various stopovers, operations which may lead to the reorganization of the load. The objective is to minimize the cost and the time spent at each stopover.
The aim of this thesis is the development of an approach for the loading of ships. This problem is difficult. This involves optimizing the number of handling operations carried out (due for example to the fact that the unloading of a particular container may require the movement of other containers which block access) as well as the time spent at stopovers , while taking into account the constraints related to the structure of the ship and the particularities of the containers (constraints imposed by standardizations and regulations). The combinatorial aspect (there are as many ways of placing n containers as there are permutations of a sequence of length n) and the fact that the number of containers is very large (between two stopovers the ship can transport up to 'to 20,000 containers), make any exact approach inapplicable.
Initially, it will be necessary to list the operational constraints and business rules provided by the industrial partner (in this case CMA CGM). A review of the state of the art will be conducted in order to analyze and evaluate the models of the problem and the solutions proposed in the literature. It will then be a question of defining a model which integrates the constraints and the business rules specific to the industrial partner and which makes it possible to express all the objectives.
In a second time, it will be a question of developing one or more approaches for the calculation of loadings. This can be achieved by calling, for example, on external solvers of the SAT, MAX-SAT or CSP type, relying on the numerous works carried out within the team in these fields, or by developing an ad hoc method. -hoc of the metaheuristic type or even graphical models. Finally, it will be possible to adapt approaches from the field of planning.%, and we are lucky to have an expert in this field on the team.
Software development tasks will be carried out during this work, the team will be assisted in technical support by a computer engineer for a large part of these questions. This will allow the research work to be concentrated as much as possible on the conceptual aspects of the work that will be carried out.
General information and application:
Location: University of Aix-Marseille, Computer Science and Systems Laboratory (COALA team), Marseille
Start date: September 2022
Remuneration: The monthly salary will be around 1600 euros.
Required profile : The person sought must hold a Master 2 or an engineering degree with solid skills in computer science (in particular in algorithms and programming), artificial intelligence, combinatorial optimization and operational research.
Deposit application : The documents required for the application are as follows: * CV (at most 3 pages), * Transcripts, results and rankings of the Master's degree or equivalent (first and second years), * Cover letter, * Letters of recommendation.
Applications must be submitted before July 14, 2022 in the form of a single pdf file sent by email to christophe[dot]gonzales[at]lis-lab[dot]fr
The CRIL (Lens Computer Science Research Lab - UMR 8188)Lab at University of Artois (Lens) is hiring a one-year postdoctoral researcher to work on the POSTCRYPTUM project, funded by the Agence Nationale de Recherche (ANR) and AID/DGA. This is a three-years project focused on algebraic cryptanalysis of public key cryptosystems, focused mainly on post-quantum schemes.
For more details on the project, please check here:
https://home.mis.u-picardie.fr/~ionica/postcryptum/Welcome.html
The ideal candidate should hold a Phd degree in Computer Science, Symbolic AI, propositional satisfiability (SAT) and beyond. Skills in one or several of the following topics will be appreciated:
· SAT solving, · Problem encodings and reformulation, · Cryptography, · Pattern mining and machine learning.
The starting date is flexible, but preferably no later than December 2022.
For more information, please contact us ([email protected], [email protected], [email protected]).
Title: Design and Development of a Virtual Platform for Animal Experimentation
Research Laboratory: Laboratoire d’Informatique de l’Université du Mans (LIUM) - Computer Science Lab of Le Mans University Technology Enhanced Learning Team
Location of the Post-Doctorate: IUT de Laval – Bâtiment CERIUM2, 52 rue des docteurs Calmette et Guérin, 53020 Laval, France
Supervisor: Lahcen Oubahssi, Associate Professor in Computer Science
Contact: Lahcen Oubahssi ([email protected])
Financing: Virtual3R Project (Virtual Reality platfoRm for animal expeRimentation)
Dates: From 1st October 2022 to 30th September 2023 (12 months)
Salary: The postdoc will earn 2849 € (gross) per month.
Keywords: Virtual Reality, Animal Experimentation, Pedagogical simulation, Virtual Learning Environments, Technoplogy Enhanced Learning, 3R.
Context of the post-doctoral subject: Virtual3R project
Research on the animal model for scientific purposes is still essential today to protect the health of citizens and animals, and to preserve the environment.
Many medical, veterinary, pharmaceutical, toxicological studies, etc, all ethically validated, use the animal model, and could not be carried out on another model. These studies are governed by legislation and regulations. European and French regulations aim to reduce the number of experiments on animals used for scientific purposes. They encourage the development of alternative methods and the use of the animal model only in the absence of other methods available to meet the object of the study.
In general, these regulations are based on the 3Rs principle, consisting of reducing, replacing and refining (improving) as much as possible the use of animals. The Virtual3R project aims to offer an alternative and complementary method based on virtual reality to reduce the number of animals used in animal experimentation training while allowing students to acquire a good understanding of basic technical procedures and gestures before experimentation with real animals. Indeed, virtual reality offers new experiences to users through ever more efficient possibilities for interaction and immersion. These possibilities are of great interest in the learning domain. TEL (Technology Enhanced Learning) research has shown great interest in virtual reality technology because of its ability to simulate a wide diversity of real-world conditions.
In this context, we are interested in VRLE (Virtual Reality Environments for Human Learning) which aim to offer learners virtual educational situations. The TEL-LIUM team has developed its expertise around the design and operationalization of learning situations in a virtual reality context (Oubahssi and Piau-Toffolon 2019) (Mahdi et al. 2019) (Mahdi 2021) (Oubahssi and Mahdi 2021). Our latest research work has provided the first solution for structuring VR-oriented educational situations in the form of reusable scenario models in different learning contexts (Mahdi 2021) (Oubahssi and Mahdi 2021). As part of this project (Virtual3R) we would like to:
• Study the question of modeling the adaptation of virtual educational situations via the reusability of virtual educational objects
• Improve the existing prototype (VEA tool) by proposing a new virtual platform for animal experimentation whose objective is to meet the new needs of the biological department teachers/ students (Laval Institute of Technology, France).
Project tasks
In the Virtual3R project context, the candidate will be particularly involved in the following tasks:
• Participating in research work on the modeling and adaptation of virtual educational situations via the reusability of virtual educational objects
• Participating in the design and development of the new Virual3R platform allowing the integration and organisation of the various educational activities related to animal experimentation
• Testing and validating the results with the biological department teachers and students (Laval Institute of Technology, France).
Candidate profile & required skills We are looking for a researcher who holds a PhD in the field of Computer Science, with experience in Technology Enhanced Learning and /Or Virtual Reality.
The candidate must:
• have defended his PhD less than 3 years before the beginning of the Postdoctorate (after 1st October 2019). • have advanced knowledge in Unity / C# development • have good writing skills • have relational and teamwork skills • be autonomous, diligent and rigorous.
The candidate is not required to speak French although it would make it easier to communicate with the local teachers during the co-design and testing phases.
How to apply?
Applicants should send their application files before August 30th 2022, to Lahcen Oubahssi ([email protected]). Application files should contain at least a full Curriculum Vitae including a complete list of publications, a cover letter indicating their research interests, achievements to date and vision for the future, as well as either support letters or the name of 2 persons that have worked with them.
https://recrutement.inria.fr/public/classic/fr/offres/2022-05143
"The aim is to design and evaluate methods for searching and querying distributed data in a SoLiD ecosystem.
The ability to perform advanced searches on large volumes of data with acceptable performance is one of the foundations for the circulation of information and the construction of social applications.
The candidate will therefore have to investigate the possible solutions allowing to build on top of the SoLiD architecture capabilities for discovering services and routes and for accessing distributed data sets, by standardizing the search and filtering capabilities of the PODs. To do this, we can rely on SPARQL traversal or decentralized query type approaches to achieve the design of a pilot architecture that also meets performance challenges, for example via cache or index systems. This would allow us to foresee the diffusion of the SoLiD ecosystem on a web scale."
Apply here: https://recrutement.inria.fr/public/classic/fr/offres/2022-05143
Fabien Gandon, Wimmics (Inria, University Côte d'Azur, CNRS, I3S, France) @fabien_gandon - http://fabien.info
I am relaying this post-doctoral offer from my colleagues at CEA-Biomaps in Orsay: Post-doctoral position on Multimodal Machine Learning for PET/CT and PET/MRI Image Reconstruction.
The position is planned for a period of 2 years from October 2022. It is part of a collaboration with the CNRS, Inserm and the University of Paris-Saclay.
The main contacts for this subject are:
- Claude Comtat – [email protected] - Florent Sureau – [email protected]
My colleagues from CEA-Biomaps in Orsay propose a post-doctoral position: Post-doctoral position on Multimodal Machine Learning for PET/CT and PET/MRI Image Reconstruction.
The duration is 2 years, starting in October 2022. This project comes from a collaboration with CNRS, Inserm and Université Paris-Saclay.
If you need further information, please feel free to contact:
- Claude Comtat – [email protected] - Florent Sureau – [email protected]
We are looking to recruit an engineer (1 year fixed-term contract, possibly 2 years) to work on the production of a glove allowing interaction by micro-gestures, as well as the production of a demonstrator showing the use of this glove in different contexts.
To apply, contact us directly by email, attaching CV and cover letter.
Do not hesitate to write to us if you have any questions and/or to distribute this offer to anyone potentially interested.
Alix Goguey and Laurence Nigay [email protected] [email protected]
Job description Your mission will be to develop a robust glove prototype allowing human-machine interaction by micro-gestures (gestures of the fingers of the hand). Initially, the mission will consist in identifying a technical solution which will be the basis of the capture. Several prototypes have already been made, each prototype exploring a different capture solution. You will have access to all of our work already carried out and will work in collaboration with a team of researchers who have acquired expertise in this field. In a second step, the mission will consist in making several copies of a glove implementing the chosen technical solution, with the objective of a level of maturity TRL 4. It is envisaged to file a patent on the realized capture system.
Main missions : - Study of possible technical solutions - Realization of a glove allowing the capture of micro-gestures - Realization of a software layer allowing the use of micro-gesture events captured by the glove - Realization of a demonstrator illustrating the use of micro-gestures in various applications.
Required skills: - Electronics - IT development - Excellent material and manufacturing skills
Profile : - You hold a Bac+5 diploma in Electronics or Computer Science (Engineering School or University) - Quality: Autonomy, ability to work in a team, appetite for research in the service of users. - All applications are welcome, regardless of age, gender identity, social or ethnic origin, sexual orientation or disability.
Remuneration: - for a beginner engineer, the monthly gross salary could be between 1985 and 2300 €
- for a more experienced engineer, the gross monthly salary could be between 2288 and 3200 €
Company : Floralis, a subsidiary of the University of Grenoble Alpes (100 p. > 10 M€ turnover) works on the development and transfer of new technologies from research laboratories and the management of industrial relations of research laboratories and in particular ensures the mission of operator of the Carnot Software and Intelligent Systems Institute (iC LSI).
As part of the project entitled "iGlove" led by the Carnot Software and Intelligent Systems Institute, we are recruiting a Design Engineer who will work within the IIHM (Human-Machine Interaction Engineering) research team of the laboratory. LIG, located on the university campus of Saint-Martin-d'Hères.
We are looking for two young doctors or doctoral students at the end of their thesis for recruitment on a permanent contract to join an AI design office specializing in industrial research.
Title: AI postdoc, image processing and computer graphics
Skills: knowledge of machine learning and deep learning, image processing and computer graphics
Position: research engineer
Qualification: young doctor or PhD student at the end of the thesis
Company: Aiway is an AI consultancy created by academics who have strong experience in AI. Aiway carries out studies in the various issues of AI (NLP, image processing, time series, empirical simulation) for a wide variety of fields: environment, chemistry, construction, administration, e-commerce, marketing. We want to invest in specific case studies which should lead us to develop products/services: medical imaging, interpretation of point clouds under construction, prediction of chemical experiments. We want to strengthen our research team by recruiting 2 young doctors or two doctoral students at the end of their thesis on permanent contracts with a taste for industrial research in our fields of activity.
Mission : the research engineers will be supervised by the scientific director to conduct studies in industrial research based on the state of the scientific art. They must have the ability to use the state of the art to solve an industrial problem. They must also have good knowledge of engineering to integrate the results of these studies into an operational context.
Profile 1 : this profile has a dominant in image processing and computer graphics with good knowledge in machine learning and engineering.
Profile 2 : this profile has a dominance in artificial intelligence, regardless of the field with good engineering knowledge.
Remuneration : €36,000 Gross + an annual bonus of €6,000 net, which corresponds to a gross equivalent of €43,200
To apply : send a CV and a cover letter by email to [email protected] with the subject “postdoc aiway 2022”
From epidemics, climate change and sustainability to seismic engineering and Industry 4.0, all the crucial challenges facing mankind today require the ability to understand and engineer ever growing reliable interdependent, complex and interconnected systems and infrastructures.If you are interested in studying complex systems, understand how to manage and mitigate risks and cascade effects this is the right program for you!
The Call for Applications for the 4 year PhD Program in Modeling and Engineering Risk and Complexity of the Scuola Superiore Meridionale (a brand new School for Advanced Studies established in Naples by the Italian Government in 2019) is now out and available at
www.ssm.unina.it/en/calls-and-news/
We are looking for bright and ambitious students from any area of Mathematics, Science and Engineering to join this exciting new program.
7 fully funded 4-year scholarships are available this year
Each scholarship includes a stipend of EUR 19,000 per year which is increased by 50% when the student is spending time abroad.
Students are expected to spend at least 12 months abroad during their PhD. Each Scholarship also includes approx EUR 4,000 for research costs/travel per year
Applicants must submit a brief scientific report (description of their MSc thesis work, CV, personal statement and reference letters) following the instructions provided at the website above by no later than 24th August 2022. Applications must be submitted using the online platform available at:
pica.cineca.it/unina/dottorato38-ssm/
Successful candidates will be announced the end of September 2022 with the course officially starting on 1stNovember 2022.
For any further information check out the PhD website at www.ssm.unina.it/en/modeling-and-engineering-risk-eng-and-complexity-merc-eng/
or contact the PhD Coordinator, Prof Mario di Bernardo, via e-mail at [email protected]
Post-Doctoral/PhD position at Telecom-Paris on Deep learning approaches for social computing
*Place of work* Telecom Paris, Palaiseau (Paris outskirt)
*Starting date* From September 2022 (but can start later)
*Context*
The PhD student/post-doctoral fellow will take part in the REVITALISE project, funded by ANR ( viRtual bEhaVioral skIlls TrAining for pubLIc SpEaking). The research activity will bring together the research topics of Prof. Chloé Clavel [Clavel] of the S2a [SSA] team at Telecom-Paris– social computing [SocComp] - and Dr. Mathieu Chollet [Chollet] from University of Glasgow – multimodal systems for social skills training, and Dr Beatrice Biancardi [Biancardi] – Social Behaviour Modelling from CESI Engineering School, Nanterre.
*Candidate profile*
As a minimum requirement, the successful candidate should have: - A master degree in one or more of the following areas: human-agent interaction, deep learning, computational linguistics, affective computing, reinforcement learning, natural language processing, speech processing - Excellent programming skills (preferably in Python) - Excellent command of English
*How to apply* The application should be formatted as **a single pdf file** and should include:
- A complete and detailed curriculum vitae - A cover letter - The contact of two referees
For the post-doctoral fellow position, additional documents are required:
- The defense and Phd reports - The contact of two referees
The pdf file should be sent to the three supervisors: Chloé Clavel, Beatrice Biancardi and Mathieu Chollet: [email protected], [email protected], [email protected]
Multimodal attention models for assessing and providing feedback on users’ public speaking ability
*Keywords* human-machine interaction, attention models, recurrent neural networks, Social Computing, natural language processing, speech processing, non-verbal behavior processing, multimodality, soft skills, public speaking
*Supervision* Chloé Clavel, Mathieu Chollet, Beatrice Biancardi
*Description* Oral communication skills are essential in many situations and have been identified as core skills of the 21st century. Technological innovations have enabled social skills training applications which hold great training potential: speakers’ behaviors can be automatically measured, and machine learning models can be trained to predict public speaking performance from these measurements and subsequently generate personalized feedback to the trainees.
The REVITALISE project proposes to study explainable machine learning models for the automatic assessment of public speaking and for automatic feedback production to public speaking trainees. In particular, the recruited intern will address the following points:
- identify relevant datasets for training public speaking and prepare them for model training
- propose and implement multimodal machine learning models for public speaking assessment and compare them to existing approaches in terms of predictive performance.
- integrate the public assessment models to produce feedback a public speaking training interface, and evaluate the usefulness and acceptability of the produced feedback in a user study
The results of the project will help to advance the state of the art in social signal processing, and will further our understanding of the performance/explainability trade-off of these models.
The compared models will include traditional machine learning models proposed in previous work [Wortwein] and sequential neural approaches (recurrent networks) that integrate attention models as a continuation of the work done in [Hemamou_a], [Hemamou_b][BenYoussef]. The feedback production interface will extend a system developed in previous work [Chollet21].
As part of the MALIN project (INclusive textbooks, https://anr.fr/Projet-ANR-21-CE38-0014 ) the CEDRIC laboratory of the CNAM in Paris and the Interdisciplinary Laboratory of Digital Sciences (LISN) in Orsay offer funding for thesis in Treatment Automatic Languages and Artificial Intelligence.
The subject focuses on the extraction and characterization of multimedia contents, applied to the case of textbooks.
The offer is detailed at the bottom of this email.
To apply, please send a CV, M1 and M2 grades and a cover letter to Camille Guinaudeau ( [email protected] ), Olivier Pons ( [email protected] ) and Caroline Huron ( [email protected] ).
Please share this information in your networks. and anyone who may be interested. Cordially Camille Guinaudeau, Olivier Pons and Caroline Huron
Extraction and enrichment of multimodal content Application to the case textbooks
Context of the thesis
The ANR MALIN project aims to make textbooks usable digital schools by children with disabilities. In indeed, the currently available digital textbooks need to be adapted to be accessible to these children. These adaptations concern both the technical aspects and educational. In most cases, the manuals are adapted from artisanal way and delivery times can be several months. These constraints do not make it possible to make efficient school inclusion for children with disabilities. The objective of the ANR MALIN project is therefore to develop technical solutions in order to automate the adaptation of textbooks digital schools to make them accessible (access, processing and interaction with the contents) to pupils in situations of disability.
The ANR project is based on a collaboration between four laboratories: LISN (Paris Saclay University), MICS (Ecole CentraleSupelec), CEDRIC (CNAM), Inserm 1284 (CRI, University of Paris). The doctoral student or the doctoral student will work in interaction with master's trainees, engineers and another doctoral student associated with the project.
Thesis topic
The first objective is to design extraction approaches automatic structure of a textbook (lessons, blocks of exercises [themselves composed of instructions, statements, examples...], memo, synthesis...) and its multimedia content (texts, images, drawings, graphs, equations, curves...) from the files provided by publishers (these are most often in pdf format). Several approaches will be considered: an approach of adaptation and enrichment of automatic document structuring systems textual (thematic segmentation, discursive segmentation) taking into account the specificity and multi-modality of the data processed and an approach based on automatic image processing aimed at identify the different blocks based on the characteristics of the image, known as "Document Layout Segmentation and Analysis” [1, 2]. Recent deep learning approaches will be tested on manually annotated datasets in order to to adapt existing models and obtain results satisfactory extraction.
The second objective is to analyze the content of each extracted block in the previous step in order to categorize them into activities educational. Thus for each exercise, it will be necessary to determine which pedagogical activity(ies) should be implemented work to achieve it. For this purpose, the doctoral student or the doctoral student will have to develop learning techniques specific innovative, supervised or not, at the meeting between the natural language processing and linguistics on the one hand and the analysis of multimedia data on the other hand [3, 4]. In this case, the visual and textual modalities will be represented in a space common to perform a multimodal classification. One of the tracks to explore consists in employing models learned on data and similar issues in French and to work on adaptation (fine-tuning) of these models from a small set of annotated data [5, 6, 7].
Skills - Master's degree in computer science or TAL with a specialization in at at least one of the following areas: - automatic language processing - machine learning - Proficiency in Python (preferred language of the project) Knowledge of the main learning libraries will be appreciated.
General informations Place of work: CEDRIC laboratory of the CNAM in Paris and Laboratory Interdisciplinary Digital Sciences (LISN) at Orsay
Contract duration: 36 months
Start of the thesis: October/November 2022
Contact: To apply, please send a CV, M1 grades and M2 and a cover letter to Camille Guinaudeau ( [email protected] ), Olivier Pons ( [email protected] ) and Caroline Huron ( [email protected] ).
Posos in partnership with the Heka team from Inria, Inserm and Paris Cité University are currently looking for a candidate for a CIFRE thesis entitled *Temporality extraction from clinical texts.*
All the details are indicated in the PDF available on this link: https://drive.google.com/file/d/1wiappXWCQ77dhIF-g2mk-Cd85UoAZ0OL/view . Please feel free to distribute to any students who may be interested.