This project aims to quantify the behaviour of captive migratory songbirds under controlled laboratory conditions, using specialized surveillance thermal cameras. The primary goal is to develop software pipelines to extract behavioural features from 24-hour videos through classic computer vision signal processing methods and/or machine learning procedures.
This project is part of a large research program in the Animal Navigation Lab that aims at understanding how animals migrate. We use migratory birds as model species and try to disentangle what are the mechanisms that cause variations in migratory phenotypes. We use field and laboratory approaches to study the endogenous migratory program and to understand how different phenotypes respond to environmental cues.
We are looking for students with coding experience in Python and an interest in image analysis, computer vision and machine learning. The project technical focus can be adapted to meet the students’ inclinations and their interests.
Duration: 15 cr (10 weeks) or 30 cr (20 weeks).
Contact information
Giuseppe Bianco, giuseppe.bianco@biol.lu.se
Susanne Ă…kesson, susanne.akesson@biol.lu.se