Marine AI Engineer
Wout
Decrop
"Building AI models for plankton & vessel detection. Python, ML, & environmental research at VLIZ."
Marine AI Engineer
"Building AI models for plankton & vessel detection. Python, ML, & environmental research at VLIZ."
About Me
I'm a marine scientist at the Flanders Marine Institute (VLIZ) in Oostende, Belgium, specializing in machine learning applications for marine ecosystem monitoring. My research focuses on developing AI-powered tools for phytoplankton classification and underwater acoustic analysis.
I work on automated image classification workflows, leveraging convolutional neural networks for biodiversity assessment in the North Sea. I'm also involved in the iMagine project, developing best practices for AI-based image analysis in plankton research and underwater sound classification.
VLIZ - Flanders Marine Institute
Aug 2025 - Present
Full-time · On-site · Ostend, Belgium
Research on innovative methods for assessing spatial ecosystem patterns and environmental characteristics using advanced imaging technologies, artificial intelligence, and automated data-analysis workflows. The project improves the quality, consistency, and scalability of marine ecosystem assessments.
Project page →VLIZ - Flanders Marine Institute
Sep 2024 - Present
Full-time · On-site · Ostend, Belgium
Development and application of deep-learning techniques for marine observation across multiple spatial scales, ranging from aerial and satellite imagery to underwater imaging systems. The project advances automated ecosystem monitoring and conservation through state-of-the-art AI methodologies.
Project page →VLIZ - Flanders Marine Institute
Feb 2023 - Present
Full-time · On-site · Ostend, Belgium
Further development and optimization of image classifiers for the FlowCytometer, ZooScan, FlowCam, and PI10 imaging platform. This includes improving classification accuracy, expanding taxonomic coverage, developing data-processing pipelines, and supporting automated biodiversity monitoring workflows.
VLIZ - Flanders Marine Institute
Feb 2023 - Aug 2025
Full-time · On-site · Ostend, Belgium
Development of deep-learning models for automated plankton image classification, with a focus on improving taxonomic identification, data quality, and the scalability of marine biodiversity monitoring through advanced machine-learning techniques and image-analysis workflows.
Project page →VLIZ - Flanders Marine Institute
Jul 2022 - Sep 2022
Internship focusing on MongoDB, updating data pipelines, machine learning techniques, and several research boat days in the North Sea.
VLIZ - Flanders Marine Institute
Aug 2021
Hands-on experience operating on a boat, working with Grafana, MySQL, sonar data, and learning the inner workings of a research organization.
Universiteit Gent
2020 - 2022
Focus on computational biology, bioinformatics, and machine learning applications in biosciences.
Aarhus University
2020 - 2021
Erasmus semester in Aarhus Denmark. Hands-on projects: image recognition and regression models. Skills: Python, Machine Learning, CNNs.
Universiteit Gent
2017 - 2020
Focused on cell and gene biotechnology, bioinformatics, and practical lab experience.
Expertise
Expertise
Research
Decrop W, Lagaisse R, Mortelmans J, Muñiz C, Heredia I, Calatrava A, Deneudt K
Lagaisse R, Dillen N, Bakeev D, Decrop W, Focke P, Mortelmans J, Muyle J, Deneudt K
Decrop W, Deneudt K, Parcerisas C, Schall E, Debusschere E
iMagine Project Consortium, including Decrop W
Events
A glimpse into the events where I have shared our findings, connected with the community, and explored the future of marine science.
A pipeline for annotating underwater acoustic recordings using AIS vessel data, producing labeled WAV segments for marine research.
View on GitHub →Deep learning model for classifying vessel distances from underwater acoustic recordings.
View on GitHub →CNN-based phytoplankton species classifier for marine ecosystem monitoring.
View on GitHub →Module using on iMagine Marketplace to run the vessel distance categorizer on the cloud.
View on GitHub →
