Deep Learning for Human and Vehicle Detection in Outdoor Environments H/F

Position offered

0.

Works

Details of the tasks :

This M2 internship is part of the CoHoMa III Challenge proposed by the Battle Lab Terre and supported by the Defense Innovation Agency, with CESI participating alongside NAE, WE Access, Conscience Robotics, and DAE.

It is evident that numerous solutions exist in the literature for human and vehicle detection in various applications such as autonomous vehicles and video surveillance. However, these approaches struggle with generalization and are highly sensitive to slight domain variations. Another critical issue is the lack of datasets contextualized to military challenges in outdoor environments, which makes model training and transfer learning challenging.

Our approach is to propose a solution based on mixed datasets, combining real scenes with synthetic objects reprojected into these scenes. The goal of this M2 internship is to assess the relevance of these datasets in deep learning for human and vehicle detection. The intern will be responsible for:

1. Creating mixed datasets (synthetic and real data).
2. Training the model using these mixed datasets.
3. Testing and evaluating the proposal’s performance on synthetic datasets.

Project Context :

This recruitment is part of the CoHoMa project (Human-Machine Collaboration), co-financed by the Normandy region and the European Union. The project aims to test, improve, and mature the technological components of a “Human-Machine Collaboration” system centered around drones and rovers. The CoHoMa III challenge, proposed by the Battle Lab Terre and supported by the Defense Innovation Agency (https://www.defense.gouv.fr/aid/actualites/battle-lab-terre-soutenu-laidorganise-troisieme-edition-du-challenge-collaboration-homme-machine), serves as one of the project’s testing grounds.

Laboratory Presentation

CESI LINEACT (UR 7527), the Digital Innovation Laboratory for Businesses and Learning in support of Territorial Competitiveness, anticipates and supports technological transformations in sectors and services related to industry and construction. CESI’s historical ties with businesses are a determining factor in its research activities, leading to a focus on applied research in partnership with industry. A human-centered approach coupled with the use of technologies, as well as regional networking and links with education, have enabled cross-disciplinary research that centers on human needs and uses, addressing technological challenges through these contributions. Its research is organized into two interdisciplinary scientific teams and two application domains:

1. Team 1, “Learning and Innovating,” is primarily focused on Cognitive Sciences, Social Sciences, Management Sciences, Education Science, and Innovation Sciences. The main scientific objectives are understanding the effects of the environment, particularly instrumented situations with technical objects (platforms, prototyping workshops, immersive systems), on learning, creativity, and innovation processes.
2. Team 2, “Engineering and Digital Tools,” is mainly focused on Digital Sciences and Engineering. Its main scientific objectives include modeling, simulation, optimization, and data analysis of cyber-physical systems. Research also covers decision-support tools and studies of humansystem interactions, especially through digital twins coupled with virtual or augmented environments.

These two teams cross and develop their research in the two application domains of Industry of the Future and City of the Future, supported by research platforms, primarily the Rouen platform dedicated to the Factory of the Future and the Nanterre platform dedicated to the Factory and Building of the Future.