Semester Project: Pitch detection for birdsong

Test and compare different pitch detection methods on birdsongs.

Keywords: pitch detection

Research context

Our group studies the development of birdsongs. Like in human speech, pitch (fundamental frequency of a signal) is a key feature we consider for analyzing birdsongs since it can represent the information contained in a harmonic signal very concisely. However, the currently widely used methods have their disadvantages. For example, the YIN algorithm [1] is not very robust and is sensitive to the tuning of hyper-parameters. An improved version of YIN, the PYIN algorithm [2] is, however, too slow to be used on a big dataset.

Semester project

There already exists a study comparing different pitch detection algorithms on human speech [3], but not on birdsong recordings. We offer a semester project for testing and comparing different pitch detection algorithms. The goal of this semester project is to explore different types of pitch detection algorithms, and apply to the birdsong recordings.

Goals

Your tasks will be to review literature, implement different methods for pitch detection on birdsong recordings, and compare them. We expect a proper documentation of your code and a well structured report. We have weekly meetings to discuss outcomes, ideas, and next steps that you are welcome to join. The semester project workload is designed for a 6-week full-time work.

Your benefits

You will learn about the various approaches taken to detect pitch, and learn how to implement them for animal vocalizations. Your work will be relevant to various fields such as signal processing, bioacoustics and neuroscience.

Your profile

We are looking for a student with a basic understanding of signal processing and programming skills (Python or Matlab). To apply please send a CV and transcript of records to one of the contacts below.

Collaborators:

Kanghwi Lee: kanlee@ini.ethz.ch

Supervisor:

Prof. Dr. Richard Hahnloser: rich@ini.ethz.ch

[1] A. De Cheveigné and H. Kawahara, ‘YIN, a fundamental frequency estimator for speech and music’, The Journal of the Acoustical Society of America, vol. 111, no. 4, pp. 1917–1930, Apr. 2002, doi: 10.1121/1.1458024.

[2] M. Mauch and S. Dixon, ‘PYIN: A fundamental frequency estimator using probabilistic threshold distributions’, in 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Florence, Italy: IEEE, May 2014, pp. 659–663. doi: 10.1109/ICASSP.2014.6853678.

[3] A. Kroon, ‘Comparing Conventional Pitch Detection Algorithms with a Neural Network Approach’, 2022, doi: 10.48550/ARXIV.2206.14357.

back