cover image
Hydro-K Détection

Hydro-K Détection

hydro-kdetection.com

1 Job

1 Employees

About the Company

Hydro-K Détection est un projet de monitoring environnemental et d’analyse de signal vidéo par intelligence artificielle. Ce projet consiste à utiliser un réseau de caméras existantes ou de déployer un dispositif de caméras spécialisées dont les images sont analysée par une intelligence artificielle, permettant la détection en temps réels de polluants et de collecter des donnés et statistiques. Cette solution répond  à un besoin de monitoring pour les autorités portuaires et autres zones aquatiques, permettant d’être alerté et de pouvoir prendre des mesures de luttes directes ou préventives plus efficientes et moins couteuses. Avec cette solution, le but est de contribuer à la préservation des écosystèmes et contribuer à limiter l’impact de l’activité humaine sur l’environnement.

Listed Jobs

Company background Company brand
Company Name
Hydro-K Détection
Job Title
Ingénieur·e Stagiaire en Analyse Vidéo & Machine Learning – Détection d’Anomalies en Milieux Aquatiques (Master 2)
Job Description
Job Title: Video Analysis & Machine Learning Intern – Anomaly Detection in Aquatic Environments (Master II) Role Summary Support the R&D team in designing, building, and deploying a real‑time video anomaly detection pipeline for aquatic monitoring systems. The role covers data preparation, model development, and software integration, with a focus on 3‑D CNNs, Vision Transformers, and semi‑supervised techniques. Expectations Master II student in Engineering, AI, Data Science, Computer Science, Signal Processing or related field. Demonstrated research mindset, curiosity, scientific rigor, and ability to work independently and collaboratively. Strong motivation to experiment and deliver actionable solutions. Key Responsibilities 1. **Data Handling** – Collect, clean, and annotate video datasets; extract features such as tracking, segmentation, optical flow; build visualization tools. 2. **Model Development** – Benchmark and refine anomaly detection methods (3‑D CNN, Vision Transformer, SloMo, LSTM), implement unsupervised/semi‑supervised approaches, tune hyper‑parameters, and conduct continuous performance evaluation. 3. **Software & Integration** – Design reversible inference pipelines for near‑real‑time deployment, integrate with backend ML services, containerize using Docker, and optimize for GPU workloads on embedded platforms. 4. **Research Collaboration** – Contribute to scientific discussions, experiment design, and documentation; collaborate closely with R&D peers and external academic partners. Required Skills - Strong foundation in Machine Learning & Deep Learning. - Computer Vision expertise: image/video processing, tracking, spatial‑temporal modeling. - Proficiency in Python and frameworks: PyTorch (preferred), TensorFlow, OpenCV, NumPy, scikit‑learn. - Understanding of video architectures: 3‑D CNN, LSTM, Transformers. - Familiarity with real‑time systems, GPU optimization, Docker. - Excellent analytical, problem‑solving, and communication skills. Required Education & Certifications - Current Master II (MSc) student in Engineering, Artificial Intelligence, Data Science, Computer Science, Signal Processing, or equivalent discipline. - No specific certifications required, but experience on complex video projects and knowledge of marine biology/ ecology is a plus.
St.-ouen, France
Hybrid
09-12-2025