Master Thesis: Phased array radio localization for behavioral monitoring of birds
Evaluate the potential of using radio localization to greatly simplify our behavioral research on songbirds. Apply machine learning on high-quality datasets with ground truth from the videos to establish a new method.
Keywords: indoor localization, radio emitter localization, software defined radio, animal tracking
Our research focuses on the information processing in biological neural systems. In our group we work with the well-studied zebra finches which are capable of vocal learning, a sensory-motor skill that is present only in a few species. To study the behavior of these birds in fine details, we record three orthogonal videos to track their movements and record their vocalizations with miniature FM-senders that are attached to each bird. This setup allows us to study the vocal learning process of juvenile birds as well as the function of the bird song in sexual selection.
The goal of this master thesis project is to explore the possibility to use the phases and amplitudes of the radio receiver in the four antennas to localize the birds. This would have a tremendous impact on our behavioral research: it would eliminate the need for the costly video recording that produces an enormous amount of data that need to be analyzed by complicated algorithms to track all the birds. With a pure radio tracking the throughput could be enhanced, the costs lowered, and the analyses simplified.
We have a large data sets with ground truth from the videos that allows to use machine learning for the mapping of the antenna signals to the position and orientation of the birds We expect from this project to judge the feasibility of radio tracking for behavioral studies and a measure of the precision of this new and promising method.
Your tasks will be literature and data review, implement a machine learning algorithm, assess the performance, and identify critical factors. We expect proper documentation of your code and data, and a useful report that could lead to a publication. We have weekly meetings to discuss outcomes, ideas, and next steps. The thesis workload is designed for 6-month full-time work.
You will learn about animal behavior and radio receiver technology and how it can be used for indoor localization. You practice general machine learning and data analysis methods and work in an interdisciplinary group in fields spanning neuroscience, bioacoustics, behavior, linguistics, and signal processing.
We are looking for a student with interests in machine learning, radio technology, signal processing and programming. To apply please send a CV and transcript of records to one of the contacts below.
“Multimodal system for recording individual-level behaviors in songbird groups”, L. Rüttimann et al. bioRxiv 2022.09.23.509166; doi: https://doi.org/10.1101/2022.09.23.509166
Linus Rüttimann: rlinus (at) ini.ethz.ch
Dr. Jörg Rychen: jrychen (at) ethz.ch
Prof. Dr. Richard Hahnloser: rich (at) ini.ethz.ch