Internship - Diffusion-based Unsupervised Joint

Le descriptif de l’offre ci-dessous est en Anglais_

**Type de contrat **:Convention de stage

**Niveau de diplôme exigé **:Bac + 4 ou équivalent

**Fonction **:Stagiaire de la recherche

**Contexte et atouts du poste**:
This master internship is part of the **REAVISE** project: “Robust and Efficient Deep Learning based Audiovisual Speech Enhancement” (2023-2026) funded by the French National Research Agency (ANR). The general objective of REAVISE is to develop a unified audio-visual speech enhancement (AVSE) framework that leverages recent methodological breakthroughs in statistical signal processing, machine learning, and deep neural networks in order to design a robust and efficient AVSE framework.
- The intern will be supervised by Mostafa Sadeghi (researcher, Inria), Romain Serizel (associate professor, University of Lorraine), as members of the MULTISPEECH team, and Xavier Alameda-Pineda (Inria Grenoble), member of the RobotLearn team. The intern will benefit from the research environment, expertise, and powerful computational resources (GPUs & CPUs) of the team.**Mission confiée**:
**Principales activités**:
Recent advances in generative modeling have seen the emergence of diffusion models as strong data distribution learners. It consists of gradually turning clean data into noise and learning a deep neural network to reverse this process so that one can generate samples of the clean data distribution starting from a pure Gaussian noise for example. This generative modeling has been successfully applied in an unsupervised way individually for the task of speech enhancement [10], and speech dereverberation [5, 6]. That is, the training no longer requires paired data as opposed to the supervised; only the clean speech data is required in training, and the enhancement or dereverberation task is performed in inference with some statistical modeling. Promising results for generalization have been found for these approaches. Utilizing diffusion models [7] also attempts to solve individually in an unsupervised way, various speech restoration but with little success particularly for the speech separation task while in a supervised way and still with diffusion model [8, 9] achieve competitive performance.

The goal of this internship will be threefold:

- Address joint speech enhancement and dereverberation tasks rather than separately with diffusion models in an unsupervised way,
- Address speech separation, with a diffusion model learned in an unsupervised way (i.e. learned only on
clean speech),
- Address the three tasks (enhancement, dereverberation, separation) with a single unsupervised framework.

**Compétences**:
**Avantages**:

- Subsidized meals
- Partial reimbursement of public transport costs
- Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
- Possibility of teleworking (after 6 months of employment) and flexible organization of working hours
- Professional equipment available (videoconferencing, loan of computer equipment, etc.)
- Social, cultural and sports events and activities
- Access to vocational training
- Social security coverage

**Rémunération**:
€ 4.35/hour

**Informations générales**:

- **Thème/Domaine**: Langue, parole et audio
Calcul Scientifique (BAP E)
- **Ville**: Villers lès Nancy
- **Centre Inria**: Centre Inria de l'Université de Lorraine
- **Date de prise de fonction souhaitée**: 2025-04-01
- **Durée de contrat**: 6 mois
- **Date limite pour postuler**: 2024-12-15

**Consignes pour postuler**:
**Sécurité défense**:
Ce poste est susceptible d’être affecté dans une zone à régime restrictif (ZRR), telle que définie dans le décret n°2011-1425 relatif à la protection du potentiel scientifique et technique de la nation (PPST). L’autorisation d’accès à une zone est délivrée par le chef d’établissement, après avis ministériel favorable, tel que défini dans l’arrêté du 03 juillet 2012, relatif à la PPST. Un avis ministériel défavorable pour un poste affecté dans une ZRR aurait pour conséquence l’annulation du recrutement.

**Politique de recrutement**:
Dans le cadre de sa politique diversité, tous les postes Inria sont accessibles aux personnes en situation de handicap.

**Contacts**:

- **Équipe Inria**: MULTISPEECH
- **Recruteur**:
**L'essentiel pour réussir**:
Prospective applicants are invited to submit their academic transcripts, a detailed curriculum vitae (CV), and, if they choose, a cover letter. The cover letter should highlight the reasons for their enthusiasm and interest in this specific project.

**A propos d'Inria**:
Inria est l’institut national de recherche dédié aux sciences et technologies du numérique. Il emploie 2600 personnes. Ses 215 équipes-projets agiles, en général communes avec des partenaires académiques, impliquent plus de 3900 scientifiques pour relever les défi