Projects

A collection of my research projects in AI, spatial ecology, and biodiversity conservation, using multimodal data including remote sensing, bioacoustics, camera traps, eDNA, and citizen science observations. All the code I produce is hosted on our institutional GitHub NINAnor

🤝 Collaboration

I'm always open to new collaborations and research opportunities. If you're interested in working together on AI, ecological data science, or biodiversity conservation, please don't hesitate to reach out!

Get in Touch

🚀 Funded Projects

Biodiversity monitoring

TABMON

Deployment of acoustic recorders across Europe to study bird diversity

Website
Machine learning

ROaR

Creating models that can generalise to other classes with not much annotations

Website
Restoration

ForPEAT

Sustainable Forest Practices and Nature Restoration on Peat Soils, using bioacoustics as a way of monitoring recovery

Website
Fish conservation

FishPath

Using turbulent eddies to create paths for safe downstream migration for salmonids and eels past hydropower intakes

Website
Biodiversity monitoring

BEAGLE

EU Horizon project extending biodiversity data collection in all modalities (eDNA, acoustic, camera traps)

Website
Remote sensing

HABLOSS

Studying nature habitat loss using remote sensing data

Website

đź’» Coding Projects

Machine Learning

DCASE 2023 Challenge

Few-shot bioacoustic event detection system for rare species monitoring using advanced deep learning techniques.

Code
Machine Learning

ecoVAD

A deep learning-powered voice activity detection algorithm specifically designed for ecological soundscape analysis.

Code
Machine Learning

Snowmobile detector

A model to detect whether snowmobiles are present in near-real time

Code
Application

BEATs trainer

A high level API for training BEATs, a powerful deep-learning acoustic classifier

Code
Application

CarbonViewer

An R Shiny application designed to calculate and visualize carbon storage in peatland areas.

Demo
Application

TABMON Dashboard

An Streamlit application designed to visualize the TABMON dataset

Demo

🎓 Student Supervision

PhD Co-supervision

Julia Wiel 2024–present

Standardisation of passive acoustic monitoring methods, including detection space modelling around acoustic recorders and development of metadata standards for large-scale PAM networks.

Master's Students

Ida Serine Bjørgo & Hedda Dahle NTNU · June 2026

Pretrained Audio Embedding Models for Sound Event Detection of Arctic Fox Vocalisations — Master's thesis in Electronics Systems Design and Innovation. Supervisor: Guillaume Dutilleux; Co-supervisor: Benjamin Cretois.

Pierre Bosman Université de Liège

Passive acoustic monitoring as a complementary approach to assess phenology in bird migration — AgroBioTech.