Research project title
Detection of benthic species on the seafloor for automated inventory (research master's)
Education level
Master (research-based)
Director/co-director
Director: Guillaume-Alexandre Bilodeau
End of display
March 31, 2026
Areas of expertise
Pattern analysis and machine intelligence
Primary sphere of excellence in research
Modeling and Artificial Intelligence
Secondary sphere(s) of excellence in research
New Frontiers in Information and Communication Technologies
Unit(s) and department(s)
Department of Computer Engineering and Software Engineering
Images and video processing laboratory (LITIV)
Conditions
Bachelor’s degree in Computer Engineering, Software Engineering, Electrical Engineering, Computer Science, or an equivalent field. A strong command of programming languages is required, primarily C++ and Python. Experience in image processing or computer vision is required, as well as familiarity with machine learning methods.
To apply, please send a cover letter, your CV, and official transcripts to gabilodeau@polymtl.ca
Detailed description
Sustainable management of marine resources in the Estuary and Gulf of St. Lawrence is facing major challenges due to increasing environmental pressures, the impact of climate change, and the proliferation of invasive species. Current fisheries management and stock assessment practices, while essential, struggle to provide a comprehensive and up-to-date picture of marine ecosystems. Traditional inventory methods, such as diving surveys or traps for benthic species like lobster, are costly, limited in precision, and poorly suited for large-scale coverage across vast geographic areas.
Recent studies have demonstrated the potential and feasibility of using marine robotics and artificial intelligence (AI) for the detection and mapping of benthic species. One such study, initially focused on green sea urchins, confirmed that this technology can provide reliable, georeferenced data while remaining non-invasive to the environment. However, the lack of an integrated tool capable of processing large volumes of data and simultaneously identifying multiple species hinders the widespread adoption of this approach for sustainable and proactive marine resource management.
The goal of this project is therefore to develop computer vision tools capable of detecting a broader variety of species, taking into account specific challenges such as individual movements and partial occlusions when organisms are hidden on the seafloor.
Financing possibility
Project funded at 22 500$/year
