Skip to main content
Weiter zur Homepage der Europäischen Kommission (öffnet in neuem Fenster)
Deutsch de
CORDIS - Forschungsergebnisse der EU
CORDIS

DeepSonic: Deep sea soundscapes from machine learning

Ziel

Acoustic recordings of ocean noise offer vital insights into marine ecosystems, revealing the impacts of human activities such as noise pollution and climate change. Large baleen whales like blue and fin whales emit low-frequency calls that travel long distances and can be detected by ocean bottom seismometers (OBS)—instruments originally deployed for seismological research. This makes OBSs a cost-effective, passive tool for long-term marine mammal and ocean noise monitoring in deep and remote areas. Traditional whale call detection methods are often inefficient in noisy environments and demand extensive manual effort. To overcome this, we developed new machine learning algorithms—carried out as part of an ERC UPFLOW spin-off project—that work in low-data training regimes to detect whale vocalisations, reverberations, and other signals within the unique UPFLOW OBS dataset, the largest of its kind in the Atlantic. We detected half million of whale calls in UPFLOW’s data in just two days, compared to the months-long timeline required by existing tools. Validated on independent OBS data, our approach accurately identifies whale calls in different oceans and can even automate the tracking of individual whales with minimal manual input. The goal of DeepSonic is to deliver a user-friendly, open-source software package for ocean noise analysis, bringing our ground-breaking algorithms to users beyond the seismology community. Initially tested with stakeholders OKEANOS (Azores) and Marine Conservation Research International (MCR), it will be integrated into PAMGuard, the go-to software platform for marine mammals passive monitoring, and later scaled to European and global monitoring observatories. Beyond whale monitoring, the software can also extend to other signals from dolphins, ships and marine infrastructure. Long-term goals include building a strong IP base and commercial pathway, targeting sectors such as offshore wind, subsea cables, and environmental regulation.

Wissenschaftliches Gebiet (EuroSciVoc)

Programm/Programme

Thema/Themen

Finanzierungsplan

HORIZON-ERC-POC - HORIZON ERC Proof of Concept Grants

Alle im Rahmen dieses Finanzierungsinstruments finanzierten Projekte anzeigen

Aufforderung zur Vorschlagseinreichung

(öffnet in neuem Fenster) ERC-2026-POC

Alle im Rahmen dieser Aufforderung zur Einreichung von Vorschlägen finanzierten Projekte anzeigen

Gastgebende Einrichtung

UNIVERSITY COLLEGE LONDON
Netto-EU-Beitrag
€ 112 517,00
Adresse
GOWER STREET
WC1E 6BT London
Vereinigtes Königreich
Region
London Inner London — West Camden and City of London
Aktivitätstyp
Higher or Secondary Education Establishments
Links
Gesamtkosten
Keine Daten

Begünstigte (2)