Pitch tracking of orca calls

We have an open MSC thesis project on pitch contour extraction from vocalizations of orcas.  
Orcas are the largest member of the dolphin family and as such are an interesting species to study animal communication because of their complex social life and coordinated hunting tactics.  
To analyze the communication of orcas is generally interesting with regards to the evolution of human language, but is also an interesting challenge per se, in terms of signal processing and information theory. The general questions are: what information do they exchange and how is it encoded in the signals? Orca calls are characterized by a frequency (pitch) contour and the higher harmonics. Some of the calls are stereotyped and a catalogue of the repertoire of each population is kept. To facilitate the analysis of the calls we want to extract the pitch contours using phase locked loops (PLL). By tracking the phases simultaneously on an array of hydrophones, we should be able to better separate overlapping calls and eventually also to track the animals in real space.  
The goal of the project is to implement an algorithm and to extract the pitch contours of a small example set of calls that were recorded with an array of hydrophones. The principle working of this new method needs to be demonstrated and the limiting factors and the difficult cases to be identified.  
We are looking for a student that wants to do pioneering work and develop a new method that has the potential to open new paths in bioacoustics research. Some experience in programming in MATLAB or Python and some knowledge in signal processing are of advantage. You will acquire expertise in audio signal processing.
Keywords: Orca, dolphin, bioacoustics, phase locked loop, pitch tracking, array signal processing, localization, source separation, Kalman filter    


If you are interested in applying for this position, please send a CV and short cover letter to Jörg Rychen (jrychen (at) ethz.ch), Kaja Wierucka (kaja.wierucka (at) gmail.com), and Richard Hahnloser (rich (at) ini.ethz.ch) 

© 2022 Institut für Neuroinformatik