Joseph Chazalon

Associate Professor

Short Bio

I received an engineer degree (2008), then a Ph.D. degree (2013), both in Computer Science, from INSA Rennes. Then I worked as research engineer at the University of La Rochelle, France, from 2013 to 2018. Since 2018, I'm a lecturer at EPITA, with both teaching and research activities.

Research

Topics

My research is focused around the following core topics:

  • Computer Vision
  • Natural Language Processing
  • Information Retrieval
  • Document Image Analysis (both historical and administrative)
  • Interactive/Assisted Information Extraction
  • Datasets, Competitions and Evaluation Protocols
  • Reproducible Research

Publications

For a full list of publications, code, datasets and other resources, you can check:

If you have any trouble downloading a paper, code or other content, please send me a message immediately.

Teaching

I teach mostly technical and applied courses, for master students. Here is a short summary of the courses I'm involved in:

  • Introduction to Machine Learning (Master)
  • Computer Vision and Pattern Recognition (Master)
  • Deep Learning applied to Computer Vision (Master)
  • Massively parallel programming on GPUs using CUDA (Master)
  • Introduction to Image Processing and Computer Vision (Bachelor)
  • Docker for ML Engineers (Master)

Publications

Une approche d’extraction de données structurées à partir d’un corpus d’archives : exemple du cadastre napoléonien du Val-de-Marne

Solenn Tual, Nathalie Abadie, Joseph Chazalon, Bertrand Duménieu, Julien Perret. Une approche d’extraction de données structurées à partir d’un corpus d’archives : exemple du cadastre napoléonien du Val-de-Marne. Table ronde Regards croisés sur les recherche, Archives Départementales du Val-de-Marne, Oct 2024, Créteil (94), France. ⟨hal-04721226⟩

An approach to extract structured data from the Napoleonic land registry maps and registers

Assia Ahmed-Abdi, Solenn Tual, Nathalie Abadie, Joseph Chazalon, Bertrand Duménieu, et al.. An approach to extract structured data from the Napoleonic land registry maps and registers. Journée scientifique GdR IASIS - Intelligence artificielle pour le patrimoine image, Oct 2024, Saint-Mandé, France. ⟨hal-04721201⟩

[Code] Code source de la webapp pour la visualisation et l'édition des annotations des annuaires du projet SoDuCo.

Joseph Chazalon, Edwin Carlinet, Bertrand Duménieu, Nathalie Abadie, Julien Perret. [Code] Code source de la webapp pour la visualisation et l'édition des annotations des annuaires du projet SoDuCo.. 2024. ⟨hal-04478595⟩

[Code] A Benchmark of Nested Named Entity Recognition Approaches in Historical Structured Documents

Solenn Tual, Nathalie Abadie, Joseph Chazalon, Bertrand Duménieu, Edwin Carlinet. [Code] A Benchmark of Nested Named Entity Recognition Approaches in Historical Structured Documents. 2024. ⟨hal-04390541⟩

An approach to extract structured information from maps and registers guided by the history of the Napoleonic cadaster

Solenn Tual, Nathalie Abadie, Joseph Chazalon, Bertrand Dumenieu, Julien Perret. An approach to extract structured information from maps and registers guided by the history of the Napoleonic cadaster. 30e Conférence internationale sur l'histoire de la cartographie (ICHC 2024), Université de Lyon, UMR 5600 - EVS (Environnement, Ville, Société) & Imago Mundi, Jul 2024, Lyon, France. ⟨hal-04637689⟩