About me

My research interests are in using machine learning for automated sensing and perception of natural environments, in particular soundscapes. I aim at developing machine learning methods which help monitoring the effects that we humans have on our ecosystems through, e.g., automated species detection and biodiversity estimation in acoustic data collected by large passive acoustic monitoring device networks in the field.

I am especially interested in annotation efficient machine listening methods for bioacoustics and ecoacoustics. I believe that there is a huge potential in using acoustic sensors and machine listening methods to quantify our natural environment. In particular to monitor animal populations.

I am also interested in developing machine listening models which are suitable for the time-frequency images that they are typically applied to. Time-frequency images, with time on one axis and frequency on the other, is a different kind of image than camera images and may require another approach. I am especially interested in the the fundamental tension between resolution in time and frequency.

Previously I have been working broadly on topics such as computer vision (fashion), privacy, time-series modelling (blood glucose and vehicle trajectories), and predictive maintanance.

I am a machine learning researcher at RISE Research Institutes of Sweden and a PhD student at the faculty of mathematical statistics at Lund University. I am part of the RISE AI deep learning group. My PhD is supervised by Olof Mogren, and my academic supervisor Maria Sandsten.