STAGE - Spatio-temporally controlled visual content generation H/F

Détail de l'offre

Informations générales

Entité de rattachement

Le CEA est un acteur majeur de la recherche, au service des citoyens, de l'économie et de l'Etat.

Il apporte des solutions concrètes à leurs besoins dans quatre domaines principaux : transition énergétique, transition numérique, technologies pour la médecine du futur, défense et sécurité sur un socle de recherche fondamentale. Le CEA s'engage depuis plus de 75 ans au service de la souveraineté scientifique, technologique et industrielle de la France et de l'Europe pour un présent et un avenir mieux maîtrisés et plus sûrs.

Implanté au cœur des territoires équipés de très grandes infrastructures de recherche, le CEA dispose d'un large éventail de partenaires académiques et industriels en France, en Europe et à l'international.

Les 20 000 collaboratrices et collaborateurs du CEA partagent trois valeurs fondamentales :

• La conscience des responsabilités
• La coopération
• La curiosité
  

Référence

2024-32994  

Description de l'unité

Based in Saclay (Essonne), the LIST is one of the two institutes of CEA Tech, the Technological Research Division of the CEA. Dedicated to intelligent digital systems, its mission is to carry out technological developments of excellence on behalf of industrial partners, in order to create value.
Within the LIST, the Laboratory of Textual and Visual Semantic Analysis (LASTI) conducts its research in the field of natural language processing and computer vision to extract, classify and generate information. The laboratory's research themes include learning with few data, trustworthiness and multimodality using discriminative and generative AI.

Description du poste

Domaine

Sciences pour l'ingénieur

Contrat

Stage

Intitulé de l'offre

STAGE - Spatio-temporally controlled visual content generation H/F

Sujet de stage

The visual representation of concepts (e.g objects) varies in space and time. For instance, new car models appear, and the same brands and models are not popular in different regions of the world. These variations are not well addressed by the current generative AI models. They are also under-researched in computer vision in general. The objective of the internship is to explore the spatio-temporal grounding of visual concepts and to propose improved generation methods along these dimensions.

Durée du contrat (en mois)

6

Description de l'offre

Objectives

The objective of the internship is to explore the spatio-temporal grounding of visual concepts and to propose improved generation methods along these dimensions. Scientifically, this problem relates to lifelong-learning and fine-grained classification, two areas in which the LASTI lab researchers have proven expertise. The internship will build on this expertise to propose an innovative approach of the topic.


After an initial step dedicated to bibliographic research, the internship objectives will be to:

  • Analyze and improve an existing spatially and temporally diversified visual dataset
  • Model the spatio-temporal variability of visual concepts in an actionable way
  • Propose methods that improve the spatio-temporal grounding of synthetic content by adapting pretrained visual generators
  • Introduce dedicated protocols for the comprehensive and sound evaluation of the contribution


The internship is designed as an introduction to research, with the aim of publishing a scientific article if the results are conclusive. This initial work could be continued in a PhD thesis, within the broader framework of AI-related topics addressed by the LASTI lab

Job-related benefits:

Joining the CEA List and the LASTI means:

  • Working in one of the most innovative research organizations in the world,  addressing societal challenges to build the world of tomorrow
  • Discovering a rich ecosystem: privileged connections between industry and academia
  • Conducting research autonomously and creatively: encouragement to promote results (scientific articles, patents, open-source codes...)
  • Work in a young and dynamic team
  • Benefit from an internal computing infrastructure with more than 300 state-of-the-art GPUs
  • Receive a monthly stipend of  ~1300€
  • Have the opportunity to continue with a PhD or as a research engineer after the internship
  • Work remotely (2 days/week),
  • Receive a 75% reimbursement on public transportation, and benefit from the “mobili-jeune” aid to reduce rent

 

Profil du candidat

Diploma: M2 or final year of engineering degree

Skills:

  • Machine learning skills (deep learning, multimedia models, generative AI…)
  • Preferred: hands-on experience with visual content generation models
  • Python proficiency in a deep learning framework (preferably PyTorch)

 

In line with CEA's commitment to integrating people with disabilities, this job is open to all.

Localisation du poste

Site

Saclay

Localisation du poste

France, Ile-de-France, Essonne (91)

Ville

Palaiseau

Critères candidat

Langues

  • Anglais (Courant)
  • Français (Intermédiaire)

Diplôme préparé

Bac+5 - Diplôme École d'ingénieurs

Formation recommandée

Master Intelligence Artificielle ou assimilé (math appli, vision, TAL...)

Possibilité de poursuite en thèse

Oui

Demandeur

Disponibilité du poste

03/02/2025