A proof of concept to make sea reliable data owners trust the data anonymization to cooperate over bigger interests. I want to develop a full anonymized data environment, from sensors to weather etc ...
The goal is to develop an automatic annotation algorithm for underwater data. We envision the following steps in the challenge:
1. Handling the video data
2. Creating audio data associated with the...
A basic yet interactive user interface coupled with an API, which can take in selected underwater images together and output a 3D reconstruction of the asset.
A minimum viable product is the ultimate goal, but if time is a huge constraint, it would just be a prototype with sufficient backup on its implementation
- Upload underwater images of coral reefs (taken via smartphones or cameras).
- Use AI-based image recognition to detect and annotate coral health indicators (e.g., bleaching, algae, ...
Over the 48 hours, we aim to develop a real-time, interactive web dashboard that connects with our AI-powered floating Seabin prototype. The platform will visualize trash detection results, sensor ...
1. A web-based dashboard showing historical analysis of ocean data; live monitoring of artificially mimicked data (to replicate as the data coming from reefSpark drone stack) and health scores of the...
We aim to prototype an interactive online platform (web dashboard) that:
- Geolocates marine species observed through EMO BON
- Enables temporal and spatial filtering of biodiversity indices
- ...
- To respond to a major social or environmental problem
- To learn and develop my skills through collaborative experience
- Contributing to innovation in the ...
We aim to develop a robust and playful algorithm that can reliably distinguish clouds from snow and ice in Sentinel-2 images of polar regions. This will include:
A trained machine learning model (or...
I want to see the feasibility and make the first prototype on a reduced scale to present the model not only of the limestone sculpture but of a functional IOT model.
Regenero proposes developing a "Marine Restoration Site-Selection & Monitoring Dashboard" prototype during the Ocean Hackathon. We would create a tool with a core decision matrix (processing ...
Develop an AI/ML driven application to classify ships based on a training dataset of hydrophone recordings with four different identified types of ships (tug, cargo, tanker, passenger). Train this ...
Develop a simple model to predict very low pH at 40m depth, based on a time series of related and co-located variables: CO2, temperature, salinity, density, oxygen, and various weather conditions. ...
An AI/ML system that can categorize plots of surface currents (measured by a coastal radar array) into 3 dominant modes, which we hypothesize to be mostly influenced by tides, river flow and ...
Campus mondial de la mer
Technopôle Brest-Iroise
525, Avenue Alexis de Rochon
29280 Plouzané Contactez-nous