Over the last years the number of insects have declined dramatically. Unfortunately, ecologists and biologists are still lacking tools for non-invasive insect monitoring. In this project the students will work with a new Wildlife Insect Dataset, which offers different visual cues such as colour information and motion cues. The aim is to (1) learn about new imaging modalities and insect datasets; (2) learn about the relevant machine learning techniques; (3) develop and train adapted machine learning models to detect the tiny insects in cluttered conditions. In particular, state-of-the-art object detection algorithms will taught, which have almost universal applicability in Geoinformatics and are not limited to the terrestrial remote sensing data presented in this course.
- Lehrende/r: Benjamin Risse