PSST! - Privacy for Smart Speech Technology  

Back

Top

“Privacy for Smart Speech Technology” (PSST) is a joint doctoral
training programme and Horizon Europe Marie Skłodowska-Curie Action, the
European Union’s flagship funding programme for doctoral training. We
are a consortium of 7 European universities and 11 industrial partners
searching for 12 PhD students to work on the protection and evaluation
of privacy for smart speech technology. PSST is a unique opportunity, as
it is the largest international project focusing on privacy in speech
technology and because the importance of privacy has only recently
gained wider appreciation.

Join us and put your expertise in deep learning / machine learning,
speech processing, information privacy and security, and user studies
into practice and gain your PhD degree from TWO leading European
Universities!

See more information and PhD topics at https://psst-doctoralnetwork.eu/

Application deadline: January 26, 2025 - apply now!

On behalf of the PSST team

Nicholas Evans
EURECOM

Back

Top

Génération de deepfakes audio-visuels par modèle de diffusion multimodal

Dates : 01/03/2025 au 31/08/2025

Laboratoire : STMS Lab (IRCAM / CNRS / Sorbonne Université et EURECOM

Lieu : IRCAM – Analyse et Synthèse des Sons (Paris) ou EURECOM (Sophia Antipolis)

Responsables : Nicolas Obin (Ircam), Jean-Luc Dugelay (EURECOM), Alexandre Libourel (EURECOM)

Contact : nicolas.obin@ircam.fr, Jean-Luc.Dugelay@eurecom.fr, Alexandre.Libourel@eurecom.fr

Contexte : Ce stage s’inscrit dans le contexte du projet DeTOX

- Lutte contre les vidéos hyper-truquées de personnalités françaises, financé par ASTRID/ANR et en collaboration avec EURECOM. Les récents challenges ont montré qu’il était extrêmement difficile de mettre au point des détecteurs universels de vidéos hyper-truquées - à l’exemple des “deep fakes” utilisés pour contrefaire l’identité d’une personne. Lorsque les détecteurs sont exposés à des vidéos générées par un algorithme nouveau, c’est-à-dire inconnu lors de la phase d’apprentissage, les performances sont encore extrêmement limitées. Pour la partie vidéo, les algorithmes examinent les images une par une, sans tenir compte de l’évolution de la dynamique faciale au cours du temps. Pour la partie vocale, la voix est générée de manière indépendante de la vidéo ; en particulier, la synchronisation audio-vidéo entre la voix et les mouvements des lèvres n’est pas prise en compte. Ceci constitue un point faible important des algorithmes de génération de vidéos hyper-truquées. Le projet DeTOX vise à implémenter et à apprendre des algorithmes de détection de deepfakes personnalisés sur des individus pour lesquels on peut disposer et/ou fabriquer de nombreuses séquences audio-vidéo réelles et falsifiées. En se basant sur des briques technologiques de base en audio et vidéo récupérées de l’état de l’art, le projet se concentrera sur la prise en compte de l’évolution temporelle des signaux audio-visuels et de leur cohérence pour la génération et la détection. Nous souhaitons ainsi démontrer qu’en utilisant simultanément l’audio et la vidéo et en se focalisant sur une personne précise lors de l’apprentissage et de la détection, il est possible de concevoir des détecteurs efficaces même face à des générateurs encore non répertoriés. De tels outils permettront de scruter et de détecter sur le web d’éventuelles vidéos hyper-truquées de personnalités françaises importantes (président de la république, journalistes, chef d’état-major des armées, …) et ce dès leur publication. Objectifs : La génération deepfakes audio-visuels repose actuellement sur l’assemblage de deepfakes audio, visuel, et de resynchronisation labiale générés séparément. Chaque modalité possède des générateurs de référence dans l’état de l’art : par exemple, LIA [1, 2] ou DeepFaceLab pour l’image, RVC [3] pour l’audio, et Wav2lip et Diff2lip [4] pour la synchronisation labiale audio-visuelle.

L’objectif de ce stage consistera à implémenter, entraîner, et évaluer un modèle de génération de deepfakes audio-visuel par diffusion multimodale à partir de générateurs existants et optimisée sur une personnalité visée.

Les contributions attendues sont :

- L’implémentation d’un post-net basé sur un modèle de diffusion à partir de flux de données asynchrones qui, à partir d’un assemblage de générateurs séparés, homogénéise et optimise le réalisme du rendu de la génération d’un deepfake audio-visuel

- La spécialisation de la génération conditionnée sur l’identité d’une personnalité, par exemple par la mise en œuvre d’un apprentissage adversarial conditionné sur la personne.

- La génération d’une base de données de deepfakes audio-visuel sur une ou plusieurs personnalités françaises.

- La mise en œuvre de protocoles d’évaluation objectif et subjectif pour l’évaluation de la qualité et du réalisme des deepfakes générés

Le stage s’appuiera en majeure partie sur les connaissances de l’équipe Analyse et Synthèse des Sons en traitement du signal de parole et en modélisation générative par réseaux de neurones, en collaboration étroite avec EURECOM pour la génération multimodale. En outre, le ou la stagiaire pourra s’appuyer sur les implémentations existantes des générateurs audio, visuel, et de synchronisation labiale déjà réalisées dans le cadre du projet DeTOX.

Compétences attendues :

● Maîtrise de l’apprentissage automatique, en particulier de l’apprentissage par réseaux de neurones, et multimodal.

● Maîtrise du traitement du signal numérique (son, image)

● Bonne maîtrise de la programmation Python et de l’environnement TensorFlow et PyTorch et du calcul distribué sur des unités GPUs

● Autonomie, travail en équipe, communication, productivité, rigueur et méthodologie.

Rémunération : Gratification selon loi en vigueur et avantages sociaux

Date limite de candidature : 20/01/2025

Bibliographie :

[1] Wang, Yaohui, Di Yang, Francois Bremond, and Antitza Dantcheva. 'LIA: Latent Image Animator.' IEEE Transactions on Pattern Analysis and Machine Intelligence (2024).

[2] Wang, Y., Yang, D., Bremond, F. and Dantcheva, A., 2022. Latent image animator: Learning to animate images via latent space navigation. In International Conference on Learning Representation (ICLR), 2022.

[3] Retrieval-based Voice Conversion. Available online: https://github.com/RVCProject/Retrieval-based-Voice-ConversionWebUI/blob/main/docs/en/README.en.md

[4] Mukhopadhyay; S. et al. Diff2Lip: Audio Conditioned Diffusion Models for LipSynchronization. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pp. 5292-5302. 2024

Back

Top

M2 Internship: Using Speech-Based AI to Study Communicative Development

Requirement: M1 in computer science Large Language Models, such as ChatGPT, have shown impressive abilities in text-based tasks. Beyond practical applications, they have also sparked scientific discussions about the nature of human language and cognitive development, including debates around Chomsky’s theories on the emergence of syntax. 1 However, these models have limitations in advancing our understanding of how children acquire language. First, they rely on vast amounts of text data for training. Children do not acquire language through exposure to written text; their language learning is grounded in speech—an inherently multimodal signal that combines linguistic and paralinguistic information such as prosody. These features are understood to play a critical role in shaping children’s communicative development. 2 Second, children are not passive learners, they actively engage in (proto-)conversational exchanges with caregivers. Through interactions, they influence their linguistic environment, creating a dynamic feedback loop that is vital for learning. 3

Recent advances in speech language modeling provide a scientific infrastructure for the study of how multimodality and interaction shape early language development. Models like Moshi 4 represent a significant step forward by processing speech directly, without first converting it into text. This approach allows an effective integration of both linguistic and paralinguistic cues. Moshi also models interactive speech communication, enabling it to listen and respond simultaneously—just as humans do. This project aims to use such speech-based models to study children’s communicative development in unprecedented ways, addressing questions about how early conversational dynamics, prosody, and meaning interact to support language acquisition and use. Beyond its scientific contributions, this work has significant societal implications. In education, it can guide the development of more engaging, low-latency e-tutoring systems. In health, it can improve the accuracy of tools for early detection of communicative disorders, such as autism, through analysis of markers like turn-taking dynamics and prosody.

The internship will focus on the Generative Spoken Language Model (dGSLM), 5 a direct precursor to Moshi. dGSLM is well-suited for an M2 internship due to its relative simplicity, while still being capable of producing significant scientific results. The main components of dGSLM include (see Figure, extracted from the original paper):

● Encoder: HuBERT, a self-supervised speech model that encodes linguistic and paralinguistic features from raw audio

● Decoder: HiFi-GAN, a vocoder for generating realistic audio.

● Model Architecture: Duplex transformer, which supports bidirectional processing of conversational dynamics. We will fine-tune dGSLM on around 150 hours of child-adult conversations from a new corpus, which includes data from 303 children aged 4 to 9 years. This fine-tuning will adapt the model to study child-directed communication. In particular, we will explore how prosody influences turn-taking dynamics, employing methods analogous to those we use to study children’s behavior in the lab.

Practicalities

The internship will be funded ~600 euros per month for a duration of 5 to 6 months. It will take place in Marseille within the TALEP research group at LIS/CNRS on the Luminy campus. The intern will collaborate with other interns from this project, as well as PhD students and researchers from the research group.

How to apply: send as soon as possible a short application letter, transcripts, and CV to abdellah.fourtassi@gmail.com

● Application deadline: December 20th, 2024

● Expected start: February 2025 6  

1 Piantadosi, S. T. (2023). Modern language models refute Chomsky’s approach to language. From fieldwork to linguistic theory: A tribute to Dan Everett, 353-414.

2 Christophe, A., Millotte, S., Bernal, S., & Lidz, J. (2008). Bootstrapping lexical and syntactic acquisition. Language and speech, 51(1-2), 61-75.

3 Murray, L., & Trevarthen, C. (1986). The infant's role in mother–infant communications. Journal of child language, 13(1), 15-29.

4 Défossez, A., Mazaré, L., Orsini, M., Royer, A., Pérez, P., Jégou, H., ... & Zeghidour, N. (2024). Moshi: a speech-text foundation model for real-time dialogue. arXiv preprint arXiv:2410.00037.

5 Nguyen, T. A., Kharitonov, E., Copet, J., Adi, Y., Hsu, W. N., Elkahky, A., ... & Dupoux, E. (2023). Generative spoken dialogue language modeling. Transactions of the Association for Computational Linguistics, 11, 250-266.

6 Ekstedt, E., & Skantze, G. (2022). How much does prosody help turn-taking? investigations using voice activity projection models. arXiv preprint arXiv:2209.05161.

Back

Top

L’équipe SAMoVA de l’IRIT à Toulouse propose plusieurs stages (M1, M2, PFE ingénieur) en 2025 autour des thématiques suivantes (liste non exhaustive) :

Back

Top

Several fully funded three-year PhD positions are opened with LIA's Speech and Language Group, at Avignon University.

One position is in Computational Linguistics, with a specialization in language translation. This position requires knowledge of several languages (and at least English and French, C1 level) and will take place in the context of FR Agorantic. The other PhD position is on the field of human-robot interactions.

Full descriptions available at:

- 'L’écart en traduction : Compréhension, gestion et traitement des écarts linguistiques et culturelles par l’intelligence artificielle' - 'The translation gap: Understanding, managing and processing linguistic and cultural gaps using artificial intelligence'
Supervision: Fabrice Lefèvre and Laurent Lombard

https://adum.fr/as/ed/voirproposition.pl?site=avignon&matricule_prop=60863

- 'Agentic LLM for pro-active multimodal human-robot interaction'
Supervision: Fabrice Lefèvre

https://adum.fr/as/ed/voirproposition.pl?site=avignon&matricule_prop=60862


Applications for both positions close on May 26. 

Back

Top

Positions Announcement

The Speech and Hearing Science Lab (SHS Lab) at the Medical University of Vienna and the Signal Processing and Speech Communication Lab (SPSC Lab) at Graz University of Technology are jointly seeking candidates for:

· 1 PhD Candidate

· 1 Postdoctoral Researcher

Both positions focus on speech processing in digital health and are expected to start on April 1st, 2025. The selected candidates will work on voice conversion for speech pathologies, with applications in (i) disease progression modeling, including treatment effect prediction, and (ii) enhancing pathological speech using a speaking aid.

Key Research Areas & Methodologies

· Speech processing

· Voice conversion

· Speech analysis & synthesis

· Speaker embeddings

· Representation learning

· Deep neural networks

· Natural language processing (NLP)

· Artificial intelligence & machine learning (AI/ML)

Qualifications

PhD Candidate

· M.Sc. degree in a relevant field (Electrical Engineering, Computer Science, Information & Computer Engineering, Electrical Engineering & Audio Engineering).

· Experience in speech processing, preferably in voice conversion.

Postdoctoral Researcher

· PhD degree in a relevant field (related to speech processing).

· Research publications in voice conversion.

General Requirements (All Candidates)

· Independent, self-motivated work ethic.

· Strong teamwork skills.

· Excellent communication abilities.

· Fluency in English (C1 level); German is an asset.

· Willingness and eligibility to work in Graz and Vienna.

· Willingness and eligibility to travel internationally, including to the USA and Asia for conferences.

About the Institutions

The SHS Lab focuses on engineering sciences for communication disorders, integrating speech signal processing, medical data science, AI/ML, medical imaging, and biomarkers. The lab is affiliated with the Department of Otorhinolaryngology and the Division of Phoniatrics-Logopedics.

MedUni Vienna is one of Europe’s leading medical universities, affiliated with Vienna General Hospital, the largest hospital in Europe. It actively advances AI and machine learning research through its newly founded “Comprehensive Centre for AI in Medicine” and is expanding with major new facilities, including the Centre for Translational Medicine (2025) and the Eric Kandel Centre for Precision Medicine (2026).

The SPSC Lab conducts research and teaching in speech processing, audio engineering, signal processing, computational intelligence, and circuits & systems modeling. It has played a key role in organizing INTERSPEECH 2019 and is leading the development of the Graduate School of Speech Language and AI Technologies within the Unite! University Alliance.

TU Graz is Austria’s oldest technical university, known for its high-impact research, student innovation, and vibrant startup ecosystem. It provides an inspiring work environment with excellent infrastructure and university support.

Diversity & Inclusion

Austrian universities are committed to increasing female representation, particularly in scientific and leadership roles. Qualified female candidates are strongly encouraged to apply. In case of equal qualifications, preference will be given to female applicants.

Compensation & Benefits

· PhD Candidate: €37,577.40 (annual gross, 75% position).

· Postdoctoral Researcher: €49,899.15 (annual gross, 75% position).

How to Apply

Send your application (including CV, motivation letter, transcript of records, at least two references, MSc/PhD thesis, and relevant publications) to:

📧 philipp.aichinger@meduniwien.ac.at and hagmueller@tugraz.at

📅 Applications will be accepted until the positions are filled.

Back

Top

We invite applications for a PhD position in neurocognition of language. The aim of the project is to better understand the interplay between language comprehension and production by studying neurocognitive mechanisms in typical and neurological adult populations. The PhD student will be co-supervised by Anahita Basirat (Lille, France) and Vitória Piai (Nijmegen, Netherlands). The planned start date is 1 October 2025. The application deadline is 28 March 2025.

We are looking for a highly motivated and talented candidate with demonstrable experience in research. The successful candidate will be based in Lille with long-duration stays in Nijmegen and will carry out research as a member of two groups:
- Language group of SCALab at the University of Lille
- Language Function and Dysfunction group at the Donders Centre for Cognition
The studies will be conducted in French and potentially Dutch. Access to an EEG system and the opportunity to acquire expertise in state-of-the-art electrophysiological techniques are provided.

The criteria for selection include:
- a Master's degree in a relevant field, such as (neuro)psychology, cognitive neuroscience, cognitive science
- a keen interest in language, patient research, and electrophysiology
- experience in the field of psychology of language or speech science as well as electrophysiology techniques is desirable, but not mandatory
- very good proficiency in written and spoken French as well as English
Applications from excellent candidates with a profile that does not fully meet all criteria will also be considered.

Applications should include:
- a CV
- a cover letter (1 page)
- a summary of previous work (1 page)
- a copy of the Master’s diploma (if available) and transcripts
- names of two referees

Applications should be sent in a single PDF file

Back

Top

A truly interesting job opportunity: 

At Radboud University in Nijmegen, NL, we will be hiring an 

Associate/Assistant Professor: Language and Speech Technology. 

Applications welcome until 16 March! 

See for more information https://lnkd.in/eJGVmAXH

Back

Top

Le poste requiert un haut niveau d’excellence scientifique en Intelligence Artificielle Générative et Analytique pour les sciences humaines et sociales et des compétences reconnues sur la création et l’utilisation des grands modèles de langage (LLM). Différents champs d’applications en sciences humaines et sociales sont privilégiés comme l’ingénierie et modélisation des connaissances et le traitement automatique de la parole/du langage. L’intérêt porté aux applications de l’Intelligence Artificielles aux sciences humaines et sociales constitue une des spécificités de l’enseignement de l’Informatique à la faculté des lettres de Sorbonne Université. Le candidat enseignera l’Informatique et l’Intelligence Artificielle dans les différentes formations de licence (sciences du langage option informatique et Intelligence Artificielle) et de master (Langue et Informatique) du département d’Informatique, Mathématiques et de Linguistique appliquées ainsi qu’en Pix (compétences numériques) pour les étudiants de la faculté des lettres.

Recherche
Le candidat sera rattaché à l’équipe de Linguistique Computationnelle du laboratoire Sens Texte Informatique Histoire (STIH, EA 4509) qui mène des recherches disciplinaires et pluridisciplinaires en Intelligence Artificielle pour les sciences humaines et sociales avec d’autres équipes du laboratoire ainsi qu’avec d’autres laboratoires de Sorbonne Université. Dans ce cadre, le candidat devra démontrer une excellente maitrise des approches pour la modélisation, l’analyse et la génération des contenus en sciences humaines et sociales notamment langagiers et interactionnels, dont les architectures des grands modèles des langage (LLM) et des RAG (Retrieval-Augmented Generation) basés sur les connaissances. Il devra avoir pris en compte dans sa recherche les aspects culturels, écologiques (IA frugale) et normatifs des recherches en Intelligence Artificielle. Il devra présenter un programme de recherche s’insérant dans ces thématiques et ces problématiques.

Recherche : claude.montacie@sorbonne-universite.fr
Enseignement : laurence.devillers@sorbonne-universite.fr
UFR : maria-victoria.eyharabide@sorbonne-universite.fr

Back

Top

La  jeune start up française 52-Herz collaborant  avec l'INRIA & L'IFREMER pour développer un appareil de communication sous-marin pour plongeur recrute  un post doc pour travailler sur le traitement de la déformation de la parole du plongeur dans l'eau. Elle dispose d' une puissance de calcu lembarquée pour tenter de faire cela  et travaille sur le débruitage mais également sur la récupération des effets de plosion. 

Voici l'offre de post-doc INRIA :

https://jobs.inria.fr/public/classic/fr/offres/2025-08624

A noter que la fin des candidatures est le 31/03.

Back

Top

Proposition de thèse financée (ANR FRENCHMELO)

Contact : Amandine Michelas [amandine.michelas@univ-amu.fr] et Sophie Dufour [sophie.dufour@univ-amu.fr]

Lieu : Laboratoire Parole et Langage (LPL, CNRS et Aix-Marseille Université)

Candidature jusqu’au 30 avril 2025 (envoyer CV).

Titre : Le bilinguisme : un atout pour le traitement de l’accentuation ?

Proposition : Il est bien connu que les francophones natifs ont des difficultés à discriminer deux mots qui diffèrent par la position de l’accent (comme les mots espagnols bebé « bébé » et bébe «il/elle boit »). L’objectif de cette thèse sera de mieux comprendre ces difficultés par le biais du bilinguisme. En particulier, nous examinerons l’impact de l’acquisition d’une langue à accent lexical soit en première langue (ex. bilingues espagnol-français) soit en langue seconde (ex. bilingues français-espagnol) sur la capacité des auditeurs à traiter des différences d’accent. D’un point de vue sociétal, cette thèse permettra de mieux comprendre les difficultés que rencontrent les francophones natifs lorsqu’ils apprennent des langues étrangères et aura ainsi des implications pour l’enseignement des langues.

Back

Top

Postdoc Position
We invite applications for a post-doc position at the Faculty of Psychology and Educational Science to
work on an SNSF research project on language production in typical young and older adults and in
patients suffering from aphasia following stroke, with behavioural and neuroimaging approaches.
Qualifications requested:
- PhD in the field(s) of cognitive psychology of language, and/or neuropsychology of language, and/or
neuroscience of language
- Solid experience with neuroimaging techniques (EEG/MEG)
- Good programing skills (Python, Matlab, R)
Starting date: September 2025 or later
Applicants should send (i) a cover letter with statement of research experience, interests, and career
plan, (ii) a CV with the names of two possible referees, and (iii) a ~300-word description of previous
research and publications by April 13 th 2025, to:
Marina Laganaro, marina.laganaro@unige.ch

Back

Top

Le Laboratoire Bordelais de Recherche en Informatique (LaBRI) ouvre un poste de MCF dans l'équipe Traitement et Analyse de Données (TAD) du département Image et Son (I&S) au LaBRI.

Pour toute information complémentaire, contacter Jean-Luc Rouas, www.labri.fr/~rouas

Back

Top

Un poste de MCF en psychologie du langage et neurocognition est ouvert à l’Université de Lille :

https://www.univ-lille.fr/fileadmin/user_upload/Universite/travailler_a_l_universite/R.H_postes_EEC/Enseignants_Chercheurs/Fiches_de_postes_Synchro_2025/16_MCF_252149.pdf
N° ODYSSEE : 252149

La personne recrutée rejoindra l’équipe Langage du laboratoire SCALab UMR 9193 (https://scalab.univ-lille.fr/laboratoire/equipes-de-recherche/langage/).

Back

Top

For my new research group “AI – Multimodal Modelling and Learning” at the University of Marburg (in collaboration with hessian.AI, the Hessian Center for Artificial Intelligence), I am offering a position (4+2 years) for a

Postdoctoral Researcher

(official job advertisement here: https://stellenangebote.uni-marburg.de/jobposting/283d40458a3570725bf80921a88ec09a44400883)

The position is offered for a period of 4 years (with the option for a 2-year extension upon successful evaluation), if no former times of qualification must be considered. The earliest starting date is April 1, 2025. The position is fulltime with salary and benefits commensurate with a public service position in the state Hesse, Germany (TV-H E 13, 100 %).

Your tasks:
- Research and development of novel AI methods in the topic areas listed below (see „Your qualification“)
- Publication of research results in high-ranked international venues (A*/A/Q1) 
- Acquisition of third-party funds for research projects (both in contributing and independent roles)
- Co-supervision of students and PhD students 
- Teaching (lectures and/or seminars)
- Optional: Setting up your own junior research group

Your qualification:
- Completed university degree (Diploma, Master‘s or equivalent) in Computer Science 
- Very good doctorate or evidence of being in the final stages of doctoral completion
- Demonstrated expertise in one or more of the following areas: computer vision, machine learning, multimodal computing, information retrieval, human-centered AI, semantic web, visual analytics
- Optional expertise in one of the following application domains: social media and disinformation, technology-enhanced learning, learning analytics, cognitive science, medical informatics, digital humanities
- Publications in internationally renowned computer science venues in at least one of the above-mentioned areas
- Excellent programming skills in common programming languages (Python, Java, etc.), experience with machine learning libraries 
- Experience in supervising student theses and collaboration in joint publications

What we offer
- Oustanding career development opportunities, e.g. towards becoming a research software engineer, mentoring and support for planning your professional career, support with grant applications
- An excellent and dynamically evolving research environment in the Department of Mathematics and Computer Science, including three newly established AI professorships
- Connection to hessian.AI (Hessian Center for Artificial Intelligence) with exceptional collaboration opportunities and high-performance computing resources for training large-scale AI models
- An excellent international and national research network (including connections to various institutes of the Leibniz Association and Fraunhofer Society)
- Funding for conference participation
- Hessian public transport ticket (Landesticket Hessen)
                
How to apply: 
Please apply here: https://stellenangebote.uni-marburg.de/en/jobposting/283d40458a3570725bf80921a88ec09a444008830/apply

Contact
If you have any questions, please write to:
Please *do not send applications via e-mail*.
Prof. Dr. Ralph Ewerth
rewerth@informatik.uni-marburg.de

Back

Top

 Research Scientist/Postdoc at the School of Computer Science at Carnegie Mellon University:

We are looking for a highly motivated and talented research scientist/postdoc candidate in multimodal human behavior modeling in real-world contexts and applications. We are looking for candidates with strong ML and CV expertise and that are excited to expand their experiences to topics related AI for Healthcare.

The ideal candidates must have a PhD in Computer Science or related fields and strong track record of publications in the top ranked ML/CV venues.

Location: Carnegie Mellon University.
Work type: Full time.
Anticipated Start Date: Now.
Position Duration: 1-2 years. Initial contract is for one year. Second year contract is based on performance.
Application: If interested, please submit a single PDF file titled FirstNameLastName.pdf, including:
 1- A brief letter of application, describing your qualifications and relevant experience to the
 position of interest (with expected date of availability),
 2- A detailed CV including a list of publications and two recent representative publications,
 3- Three reference letters (sent separately by the referees).

Please visit the job details page for more information and submit the one single PDF file with all requested information (Points 1-3):
https://cmu.wd5.myworkdayjobs.com/en-US/CMU/job/ROB---HAMMAL---Postdoctoral-Fellow_2022833

Thank you
Zakia Hammal

Back

Top

Inria, the French national institute for research in digital science and
technology, is opening a fully-funded PhD position on differential
diagnosis of heart attack from speech:
https://jobs.inria.fr/public/classic/en/offres/2025-08716 (deadline:
April 30).

Candidates shall apply online at their earliest convenience.
Applications will be assessed on a rolling basis.

Back

Top

3 PhD positions in speech deepfake detection and automatic speaker verification at EURECOM

The Audio Security and Privacy Group at EURECOM, France has openings for 3 PhD candidates in speech deepfake detection and automatic speaker verification (ASV). If you have a Master's degree, an excellent academic track record, strong proficiency in English, also have expertise in computer science, machine learning, artificial intelligence, data science, speech processing, deepfake/spoofing detection, text-to-speech synthesis or voice conversion, and are keen on international collaboration, please consider applying.

Topics include, but are not limited to:
- Integrated solutions to spoofing robust ASV
- Trojan back-door attacks against ASV systems
- Watermarking of synthetic speech and converted voice
- Source tracing for spoofing/deepfake attacks
- Adversarial attacks
- Audio-visual deepfake detection

For these particular PhD positions, applications may undergo administration security checks in compliance with French law and regulations. Restrictions on certain nationalities may apply.

In the first instance, please send your CV by email to Nicholas Evans (evans@eurecom.fr) with the subject line 'PhD opportunities'.

Learn about EURECOM by visiting our website https://www.eurecom.fr and about other job opportunities at https://www.eurecom.fr/en/eurecom/job-opportunities/job-opportunities. 

Back

Top

Nous avons ouvert à Leiden(LUCL) un poste de doctorant.e en linguistique

 française quicommencera en septembre 2025, pour travailler sur la

phonologie/phonétique de créoles à base française. Avec Jenny
Doetjes, nous cherchons des candidat.es francophones, avec un master
en linguistique, idéalement en phonologie/phonétique, et qui seraient
aussi capables d'enseigner dans le département de français (langue,
linguistique). Est-ce que vous pensez que cela pourrait intéresser
vos étudiant.es ? La date limite pour postuler est le 31 mars (donc
très bientôt), mais nous allons demander une extension donc il
devrait y avoir un peu plus de temps. L'annonce se trouve ici :
https://www.universiteitleiden.nl/vacatures/2025-nl/q1/15526phd-candidate-in-french-linguistic

Back

Top

The Institute for Intelligent Systems and Robotics at Sorbonne University (Paris) is looking for a highly motivated and ambitious engineer or postdoctoral researcher to conduct research in machine learning for human-robot collaboration.

Context and objectives

This position focuses on developing machine learning techniques to enhance human awareness in human-robot interaction by integrating situation assessment and action planning.
The successful candidate will contribute to cutting-edge research at the intersection of robotics, artificial intelligence, and human interaction, with an emphasis on designing and evaluating robotic systems that facilitate seamless collaboration with humans.

The position is for 18 months contract, but there is a possibility to be extended depending on the performance and circumstances. 
The position is open at both the engineer and post-doctoral levels for candidates with a strong background in machine learning, human-machine interaction, or robotics.

Responsibilities:
    • Develop advanced situation assessment techniques using machine learning to accurately represent user preferences, behaviors, and characteristics based multimodal data to efficiently plan actions.
    • Collaborate with interdisciplinary teams, including computer scientists, experts from the humanities, and designers, to ensure the usability and effectiveness of the developed techniques.
    • Publish research findings in top-tier conferences and journals in the field of Human-Machine Interaction and Machine Learning (mainly at the post-doc level)


Requirements   
The successful candidate should have: 
    • Experience in human-machine interaction
    • Good knowledge of Machine Learning Techniques
    • Good knowledge of experimental design and statistics 
    • Excellent publication record
    • Strong skills in Python 
    • Willing to work in multi-disciplinary and international teams
    • Good communication skills

Application 
Interested candidates should submit the following by email in a single PDF file to: mohamed.chetouani[@]sorbonne-universite.fr with the subject: Application ML for Human-Robot Collaboration

    • Curriculum vitae with 2 references (recommendation letters are also welcome) 
    • One-page summary of research background and interests 
    • At least three papers (either published, accepted for publication, or pre-prints) demonstrating expertise in one or more of the areas mentioned above 
    • Doctoral dissertation abstract and the expected date of graduation for a post-doc position levale (for those who are currently pursuing a Ph.D) 

Application’s deadline: April 21, 2025.

Back

Top

We are offering a PhD position on the topic: 'Automatic Extraction and Structuring of Cultural Events: An Efficient and Frugal Approach to Connect ICC Stakeholders and Their Audiences.'

Location: Laboratoire Informatique d’Avignon (LIA) or Laboratoire des Sciences du Numérique de Nantes (LS2N)
Start Date: Fall 2025
Funding: 3 years (subject to approval by PEPR ICCARE, ~€1700-1900/month net)
Supervisors: Vincent Labatut vincent.labatut@univ-avignon.fr (LIA), Richard Dufour richard.dufour@univ-nantes.fr (LS2N)
Industrial Partner: Thomas Chenevier thomas@ideactiv.com, ideactiv (https://www.ideactiv.com)

Objective
Develop an automatic system to extract and structure cultural event information from the websites of sector stakeholders (festivals, museums, performance venues, etc.). The approach should be efficient and frugal, minimizing errors and hallucinations from language models.

Research Areas
- Information Extraction & NLP (Named Entity Recognition, Disambiguation, etc.)
- Analysis of Unstructured Web Content
- Machine Learning Models

Candidate Profile
Master’s degree or engineering diploma in computer science (or equivalent), with experience in NLP, machine learning, and/or software engineering. Strong English proficiency, autonomy, and teamwork skills are essential.

Application Process
Send your CV, academic transcripts, recommendation letters, and a motivation letter (specific to the topic) to the supervisors listed above.

More information and detailed document here: https://uncloud.univ-nantes.fr/index.php/s/WagAfz8f3MSy6Zj

Back

Top

In the context of the ENACT AI cluster, a funded three-year PhD position will open with LORIA's MultiSpeech Team (Université de Lorraine) and LIA's Speech and Language Group (Avignon University):

'Social-behavior-aware chatbot for a communication skills coaching of medical students'
Supervision: Irina Illina, Patrice Gallet and Fabrice Lefèvre

Full description and application form available at https://doctorat.univ-lorraine.fr/fr/les-ecoles-doctorales/iaem/offres-de-these/enact-chatbot-axe-sur-le-comportement-social-pour-le

Applications should include (in PDF):
- cover letter (1 page max)
- resume
- summary of previous work (1 page)
- Master’s diploma (if available) and transcripts for all years
- two referees to contact

Applications close on April 24. 

Back

Top

Nous proposons une offre de thèse en informatique sur le sujet « ANALYSES À BASE D'APPROCHES NEURONALES DE LA PAROLE DÉGRADÉE DANS LE CONTEXTE DE TROUBLES DE LA PAROLE, EN VUE DE SA RESTAURATION  » co-dirigée par Corinne Fredouille (LIA) et Julie Mauclair (IRIT)

Lieux : Laboratoire Informatique d’Avignon (LIA) et/ou Institut de Recherche en Informatique de Toulouse (IRIT)
Début : 1e octobre 2025

Back

Top

Background:

It is known that many modalities (e.g. articulation, mouth movements, eye gaze, head nods, back channels and gestures) play a role in communication success in speech-based interaction. Lombard speech is one way we modify the sound of our voices to ensure we communicate clearly with others in noisy environments, but recent research has shown that Lombard speech in reality is multimodal. This PhD student will focus on approaches to tracking a conversation that incorporate knowledge from the visual modality of speech to support understanding in noisy conditions. By gaining deeper insights into how multimodal cues are exploited in noisy conditions, the research will look at approaches to architectures that approach the level of flexibility humans have in adapting to changing communication challenges. This PhD research is part of a larger SFI Frontiers project SpeechSpace focussed on multimodal speech and led by Prof. Harte. The funding available will cover fees and stipend, but also equipment and conference travel associated with the research.

Person Requirements:

• Primary or Master’s degree in Electronic Engineering or closely related discipline, with an interest in engaging in multidisciplinary research
• Must meet TCD University requirements at https://www.tcd.ie/study/apply/admission-requirements/postgraduate/
• Strong skills in coding, machine learning, deep learning (basics is fine), and signal processing with the willingness and motivation to learn new skills and packages
• Prior experience with speech-based interaction desirable
• Excellent communication skills, both spoken and written, and fluency in English

The position:

• Prof Harte’s research group brings together a diverse and friendly group of people who are all interested in pursuing research in speech-related topics and like to share ideas and learn from one another. As such, this is a fully in-person position for someone who wants an in-person PhD experience and would like to contribute to our group both technically and socially.

To apply for the position, please complete the following application form:

https://forms.gle/YAif8DfV7eSVRdxM8

Informal inquiries only to Prof Naomi Harte at nharte@tcd.ie

Back

Top

=== Workshop SIR ===

Atelier sur les Avancées en AMR et en Analyse Sémantiques
SIR@IXCS2025 - Düsseldorf - September 24 2025

=================================

	https://team.inria.fr/semagramme/first-workshop-on-semantics-for-interdisciplinary-research/

	https://openreview.net/group?id=inria.fr/INRIA/S%C3%A9magramme/2025/SIR01

=================================

In recent years, Natural Language Processing (NLP) has increasingly intersected with the humanities and social sciences, 

offering new methodologies for analyzing textual data, interpreting meaning, and modelling (IF WE WANT BRITISH SPELLING, 

WE MIGHT NOT?) language-based phenomena. The potential for multi-disciplinary research using NLP methods is particularly  

 great in computational semantics (CS, as its ability to process and represent meaning opens up innovative pathways for 

researchers in history, philosophy, literary studies, political science, etc.  This workshop aims to explore how semantic

 models and tools can be leveraged to tackle traditional and emerging questions in the Humanities in a broader sense (Social 

Sciences, Law, Economics, Management, Literature, Languages, Art, …). 

A major theme of  SIR is the role of semantics in NLP applied to the humanities (both statistical and symbolic approaches).

=== Topics to Explore ===
    • CS and the humanities: issues, tools and applications.
    • Quantitative and qualitative approaches as a breakthrough in the  Humanities
    • NLP transforming humanities issues
    • Contributions and limitations for understanding meaning
    • Links between formal semantics and neural models
    • Ambiguity, polyphony and interpretation in theHumanities
    • Ethics and bias in semantic modeling
    • Interdisciplinary dialogue between AI, NLP and Humanities
 
=== Dates ===
    • Deadline : July 14th (anywhere on earth)
    • Notification : August 25th (anywhere on earth)
    • Camera Ready : September 10th (anywhere on earth)
    • Workshop : September 24th (anywhere on earth)
 
=== Submission Information ===
Papers should describe original research and must not exceed 4 pages (with an extra page in the camera ready version for accepted papers). Papers should be submitted no later than 14 July 2025 (anywhere on earth).
 
Accepted papers will be published in the conference proceedings in the ACL Anthology. For inclusion in the proceedings, at least one author must register to the conference and present the paper in person. 
 
Submissions should be fully anonymous to ensure double-blind reviewing.
 
=== Submission ===
https://openreview.net/group?id=inria.fr/INRIA/S%C3%A9magramme/2025/SIR01
 
=== Style Files ===
The workshop follow the IWCS 2025 template see the workshop web page.
 
=== Organizers ===
Maxime Amblard, Université de Lorraine
Ellen Breitholtz, Gothenburg University

=== Contact ===
maxime.amblard@univ-lorraine.fr and ellen.breitholtz@ling.gu.se

Back

Top

Bonjour à tous, 

Back

Top

Le GIPSA Lab à Grenoble (France) propose un sujet de thèse portant sur la 'Reconnaissance automatique du Human Beatbox et application à la compréhension d’un langage musical évolutif'.
Pour toute info et pour candidater : https://adum.fr/as/ed/voirproposition.pl?langue=&matricule_prop=65500&site=edisce

Back

Top

We are very happy to announce an opening for a
*** Tenure-Track Position in Natural Language Processing and Interaction *** at the Johannes Kepler University, Institute of Computational Perception, Linz/Vienna, Austria. The position is open to female applicants only.

More details can be found at https://www.jku.at/en/the-jku/work-at-the-jku/job-openings/tenure-track-positions/natural-language-processing-and-interaction/

Key facts:
- Full-time employment (40h/week)
- Six-year tenure-track position as Assistant Professor (gross salary ~69K €)
- After successful completion of a qualification agreement, resulting in a permanent position as Associate Professor (gross salary ~81K €)
- Research and teaching duties
- PhD degree in Computer Science, Artificial Intelligence, Machine Learning, or related discipline required
- Research focus in a current field of NLP or LLMs, both in basic research (e.g., new architectures for LLMs) and applied research (e.g., misinformation detection, recommender systems, social media mining).
- Open solely to female applicants (in accordance with JKU's Development Plan)
- Application deadline: May 28, 2025
- Earliest starting date: Oct 1, 2025 

Back

Top

Nous proposons un sujet de thèse sur l’étude des processus d’adaptation et de prise de décision chez des personnes qui bégaient.

Plus d’infos sur le sujet et les modalités de candidature sur le site https://adum.fr/as/ed/voirproposition.pl?langue=&site=edisce&matricule_prop=65236

Back

Top

Profile Required

Holders of a doctorate or a PhD or equivalent degree or applicants who have gained scientific qualifications or carried out scientific work deemed to be of an equivalent level.There is no restriction on the age or nationality of applicants. All CNRS positions are accessible to people with disabilities, with special arrangements for tests made necessary by the nature of the disability

Institution Strategy

Digital health is one of the French government's strategic priorities. This has led to the funding of a Priority Research and Equipment Programme (PEPR) for digital health, but also in the inclusion of computer science issues in various national health programs that have already been launched or are in the process of being brought to full development for example in the fields of mental health (PEPR ProPsy, co-supervised by the CNRS and Inserm), neurosciences and prevention. All these initiatives are highly interdisciplinary, involving close collaboration between researchers in life sciences and computer sciences, in interaction with clinicians.
These areas are also in line with the challenges identified by the CNRS in the cross-disciplinary “Brain” and “AI foundations” areas of its next priority objectives and resources contract (CNRS COMP).
For the CNRS, one of the challenges is to develop the use in its laboratories of the multiple potentialities offered by generative AI in the fields of biology and health, and to develop new fields of research in the computer sciences whose application developments will be adapted in particular to health-related use cases.

Host Lab Strategy

The CPJ GenAI-Psy aims to strengthen an interdisciplinary multi-team research project (PRIME) at the interface between UMR6033 SANPSY (CNRS-UB) and UMR5800 LABRI (CNRS-UB). The LABRI is a leading institute in artificial intelligence and digital health through two cross-disciplinary axes that bring together more than 80 researchers. SANPSY focuses specifically on sleep, sleepiness, addiction and neuropsychiatric diseases, and how they can be monitored using digital tools. For several years now, LABRI and SANPSY have been working closely together to develop conversational agents for health. The aim is to jointly develop multi-agent approaches based on generative AI and large language models (LLM), apply them to medical problems, and validate them in use cases such as the creation of analysis tools to help diagnose and manage pathologies, and the design of predictive models of their evolution.

International Strategy

The successful candidate will be expected to implement actions and partnerships in the thematic field of the Chair, and in particular to submit a project in response to European calls for proposals (ERC, Horizon Europe Consortium, European Partnership Brain Health, etc.) during the first 4 years of his/her contract.

National Directory of Research Structures (RNSR) of the host laboratory

SANPSY - Sommeil, Addiction et Neuropsychiatrie - (UMR6033)201119462L
LaBRI Laboratoire Bordelais de Recherche en Informatique - (UMR5800)199511665F

Summary of the scientific project

The project will be part of the joint LABRI-PRIME and SANPSY team. The aim is to develop and validate multi-agent approaches based on large language models (LLMs), applied to medical problems such as the creation of diagnostic tools based on linguistic data and pathology management, the generation of interactive scenarios run by conversational agents, the analysis of health data corpus and the design of predictive models for the evolution of disorders. One challenge will be to ensure the robustness of LLMs to enable their use in complex and unpredictable human behavior disorders, to ensure the reliability of models in clinical environments, and to participate in their validation with patients. The project of the laureate, while firmly embedded in computer science, will therefore have to be based on use cases developed in collaboration with the site's clinicians.

Summary of the teaching project

The successful candidate will join the teaching teams at the University of Bordeaux. On the one hand, he/she will be able to contribute to the teaching of artificial intelligence, in particular its application to health and neuroscience. He/she will also be able to contribute to medical training on textual data analysis in healthcare and training in Neuropsychiatry.

Financial Environment

• Total (included package ANR) : 200 k€
• Total financement : k€

Scientific dissemination

The dissemination of the results will be done through world-class scientific productions: publications, patents, software... In addition, the results will be communicated to various targets such as scientific communities, media, decision makers, general public, schools, etc., with an adapted calendar. Specific tools may be developed such as websites, newsletters, meetings, international symposia, summer schools and conferences.

Open Science

The CNRS is developing a strong policy in favor of open science. Open science consists of making research results 'as accessible as possible and closed as necessary'. As such, the CNRS aims to make 100% of the texts of publications resulting from the work of its laboratories accessible , in particular through deposit in HAL. The data produced must also be made available and reusable, except for specific restrictions. In addition, the guiding principles of individual evaluation have been revised in accordance with the DORA declaration, to be more qualitative and to take into account all facets of the researcher's profession.

Science and society

The relationship between science and society is now recognized as a full dimension of scientific activity. The project will develop this dimension in synergy with all the partners. The resulting research work will contribute to informing public decision-making. Participatory science initiatives may be initiated with actors from the project’s socio-economic and cultural eco-system.

Indicators

The activity will be evaluated in particular on the basis of scientific production (publications, software, patents, etc.), on institutional and private partnerships formalized by contracts, on international presence, on the promotion of work to multidisciplinary scientific communities, on innovation and its transfer to society and on scientific dissemination to non-specialist audiences.

Methods of organization of the interviews

Only the candidates selected on file by the selection committee will be invited to the auditions.

Back

Top

we offer a PhD position in multimodal automatic detection of stuttering-like disfluencies.

Workplace: Nancy, Montpellier

Start date: October 1st 2025

Required Skills:
The candidate should hold a Master's degree in computer science, have strong skills in machine learning and deep learning, and be proficient in Python and frameworks such as PyTorch or TensorFlow. An interest in signal processing (audio/video) and ideally in NLP is expected. Autonomy, rigor, critical thinking, and analytical abilities are essential, along with strong communication skills to work in a multidisciplinary environment. An interest in phonetics, linguistics, and speech disorders—particularly stuttering—would be a plus.

For full details and application instructions:

Back

Top

We offer a one-year pos doc position in our SpeechStab project. Details is found at the link below.

https://emploi.cnrs.fr/Offres/CDD/UMR5216-CHRROM-035/Default.aspx?lang=EN <https://emploi.cnrs.fr/Offres/

CDD/UMR5216-CHRROM-035/Default.aspx?lang=FR>

Required skill can be having scientific background in neuroscience and/or experimental psychology and/or human 

motor control, who has substantial computer programming experience (in particular MATLAB, or Python, or C language), 

and to be motivated in experimental work with human participants and/or applying computer models of motor control to 

interpret experimental data in perturbed speech production. Applicants must hold a PhD degree (or be about to earn one).

 A field of computer science or bioengineering is desirable, but not mandatory.

Our offer is one post, but we prepare two possible subjects for this post. A candidate can select either subjects or both. 

The required skill can be depending on the subjects of interests. 

Back

Top

Offre de thèse en informatique sur le sujet « EXTRACTION D'INFORMATIONS SÉMANTIQUES À DES FINS ÉDUCATIVES DANS 

DES RÉSUMÉS ORAUX PRODUITS PAR DES ENFANTS  » dirigée par Nathalie Camelin.

Lieux : Laboratoire Informatique d’Avignon (LIA) 
Début : 1er octobre 2025

Plus de détails et candidature ici : https://adum.fr/as/ed/voirproposition.pl?site=adumR&matricule_prop=66228

Back

Top