Postdoc: Analysis of gendered cinematography with AI

4 Août 2023

---

Catégorie : Post-doctorant

---

Mission

The postdoctoral researcher will be recruited in the Signal Image and Systems department of the I3S laboratory, in the framework of the TRACTIVE project centered on film analysis with AI (machine learing - ML) models. The TRACTIVE project is a four-year research project funded by the French National Research Agency (ANR) and involving 6 academic laboratories, in computer science and humanities. The scientific objective of TRACTIVE is to characterize and quantify gender representation and objectification in films and visual media by designing an AI-driven multimodal (visual and textual) discourse analysis. To address this objective, TRACTIVE brings together researchers from computer science, media studies, linguistics, and gender studies. The TRACTIVE project already involves two engineers and two PhD students. This postdoc proposal is at the center of the TRACTIVE project, with the following objective.

Objective: Designing explainable deep learning (DL) approaches to identify complex multimodal film editing patterns producing on-screen objectification.

More info at: https://www.i3s.unice.fr/~sassatelli/openings.html

 

Activities

Main activities are carrying research on the above subject, specifically:

• design of explainable ML models to identify cinematographic patterns
• establishing processes to manage and study multimodal data (image, text, audio, metadata)
• rigorous analysis and comparison of results
• writing scientific articles targeting most prestigious conferences and journals of the domain (CVPR, ICCV, IJCV, etc.)
• writing deliverables for the ANR funding agency

 

Additional activities

• participating in project meetings
• collaborative work with PhD students, engineers and partners working on the project

 

Pre-requisites

 

Required

• Strong research and publication experience
• Excellent programming skills in Python,
• Solid experience with machine and deep learning libraries (tensorflow, torch, scikit learn) and main models (CNN, RNN, Transformers),
• Strong competence and prior experience with computer vision and image processing tools and libraries (e.g., opencv, ffmpeg),
• Good engineering skills to debug and compile repositories with source code in C++ and Matlab,
• Sound methodology for scientific research, understanding and writing

 

Appreciated

• Experience with time series analysis
• Good familiarity with data processing workflows for machine learning,
• Expertise in code repository management (git),
• A good level of written and spoken French and English

 

Work environment

The postdoctoral researcher will be recruited in the Signal Image and Systems department of the I3S laboratory, in the framework of the TRACTIVE project centered on film analysis with AI (machine learing - ML) models. This postdoc proposal is at the center of the TRACTIVE project. In addition to the permanent researchers, several non-permanent staff members are already working on the project, including PhD students and engineer.

 

Supervisors:

Prof. Lucile Sassatelli, Full Professor, Fellow of IUF

lucile.sassatelli@univ-cotedazur.fr

https://www.i3s.univ-cotedazur.fr/~sassatelli

https://www.i3s.univ-cotedazur.fr/TRACTIVE/

Dr. Hui-Yin Wu, Research Scientist at Inria center of Université Côte d’Azur

hui-yin.wu@inria.fr , https://www-sop.inria.fr/members/Hui-Yin.Wu/

 

Remuneration

Gross salary: from 2800€ /month depending on experience

French social security included