An Automated Pipeline for 3D Flight Path Reconstruction from Multi-Channel Audio of Echolocating Bats
(2026) In Master's Theses in Mathematical Sciences FMAM05 20261Mathematics (Faculty of Engineering)
- Abstract
- This thesis presents an automated pipeline developed for reconstructing the 3D flight trajectories of bats navigating a cave environment purely from multi-channel audio recordings. The pipeline was developed and validated using the Ushichka dataset which contains multiple simultaneous audio and video recordings of bats in their natural habitat, the Orlova Chuka cave in Bulgaria. The proposed pipeline integrates computer vision, signal processing algorithms, and robust multilateration solvers. Echolocation calls are visually detected from audio spectrograms using a trained SqueezeNet Convolutional Neural Network. Time Differences of Arrival across the microphone array are then estimated via Generalized Cross-Correlation with Phase... (More)
- This thesis presents an automated pipeline developed for reconstructing the 3D flight trajectories of bats navigating a cave environment purely from multi-channel audio recordings. The pipeline was developed and validated using the Ushichka dataset which contains multiple simultaneous audio and video recordings of bats in their natural habitat, the Orlova Chuka cave in Bulgaria. The proposed pipeline integrates computer vision, signal processing algorithms, and robust multilateration solvers. Echolocation calls are visually detected from audio spectrograms using a trained SqueezeNet Convolutional Neural Network. Time Differences of Arrival across the microphone array are then estimated via Generalized Cross-Correlation with Phase Transform. To robustly estimate 3D call positions and timing in the presence of anomalous data, a minimal multilateration solver is coupled with a Random Sample Consensus framework. Finally, a multi-hypothesis tracking method associates these discrete coordinates into continuous trajectories. Processing of the acoustic data successfully yielded numerous reconstructed trajectories. Validating the acoustic reconstruction against ground-truth trajectories constructed from thermal cameras, the mean absolute orthogonal distance between them is 3.6 cm. This result demonstrates the usefulness of the approach in studying the flight of bats, and, by extension, their group behavior. (Less)
- Popular Abstract
- Eavesdropping on Bats: How Sound Reveals Their Hidden Flight Paths.
In order to navigate in the dark, bats emit high-pitched echolocation calls. Usually, no one is around to listen, but if we place microphones in a cave, we can record these calls and actually pinpoint exactly where the bats are flying. We do this by first measuring the time that a call arrives at each microphone. Because sound travels at a specific speed, the time it takes for a call to reach a microphone reveals exactly how far away the bat is. When we combine the measurements of at least four different microphones, we can solve for the exact location of the bat. This method of figuring out the location of a sound source using the inferred distance to a collection of... (More) - Eavesdropping on Bats: How Sound Reveals Their Hidden Flight Paths.
In order to navigate in the dark, bats emit high-pitched echolocation calls. Usually, no one is around to listen, but if we place microphones in a cave, we can record these calls and actually pinpoint exactly where the bats are flying. We do this by first measuring the time that a call arrives at each microphone. Because sound travels at a specific speed, the time it takes for a call to reach a microphone reveals exactly how far away the bat is. When we combine the measurements of at least four different microphones, we can solve for the exact location of the bat. This method of figuring out the location of a sound source using the inferred distance to a collection of microphones is called Multilateration.
In this project, I developed a fully automated method to robustly solve the multilateration problem. The method synergizes many different techniques, ranging from using artificial intelligence to detect bat calls in the audio recordings, to advanced mathematical algorithms. My method successfully tracked a larger number of bats with higher precision than previous attempts, allowing us to map their tangled flight paths. The ultimate goal of tracking bats in this way is to understand how they behave and communicate in groups, and this project brings us one step closer to uncovering that mystery. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/student-papers/record/9234971
- author
- Brasar, Sparf Nils LU
- supervisor
-
- Karl Åström LU
- organization
- alternative title
- En automatiserad pipeline för 3D-rekonstruktion av flygvägar baserat på flerkanalsljud från ekolokaliserande fladdermöss
- course
- FMAM05 20261
- year
- 2026
- type
- H2 - Master's Degree (Two Years)
- subject
- publication/series
- Master's Theses in Mathematical Sciences
- report number
- LUTFMA-3629-2026
- ISSN
- 1404-6342
- other publication id
- 2026:E56
- language
- English
- id
- 9234971
- date added to LUP
- 2026-06-17 15:58:09
- date last changed
- 2026-06-17 15:58:09
@misc{9234971,
abstract = {{This thesis presents an automated pipeline developed for reconstructing the 3D flight trajectories of bats navigating a cave environment purely from multi-channel audio recordings. The pipeline was developed and validated using the Ushichka dataset which contains multiple simultaneous audio and video recordings of bats in their natural habitat, the Orlova Chuka cave in Bulgaria. The proposed pipeline integrates computer vision, signal processing algorithms, and robust multilateration solvers. Echolocation calls are visually detected from audio spectrograms using a trained SqueezeNet Convolutional Neural Network. Time Differences of Arrival across the microphone array are then estimated via Generalized Cross-Correlation with Phase Transform. To robustly estimate 3D call positions and timing in the presence of anomalous data, a minimal multilateration solver is coupled with a Random Sample Consensus framework. Finally, a multi-hypothesis tracking method associates these discrete coordinates into continuous trajectories. Processing of the acoustic data successfully yielded numerous reconstructed trajectories. Validating the acoustic reconstruction against ground-truth trajectories constructed from thermal cameras, the mean absolute orthogonal distance between them is 3.6 cm. This result demonstrates the usefulness of the approach in studying the flight of bats, and, by extension, their group behavior.}},
author = {{Brasar, Sparf Nils}},
issn = {{1404-6342}},
language = {{eng}},
note = {{Student Paper}},
series = {{Master's Theses in Mathematical Sciences}},
title = {{An Automated Pipeline for 3D Flight Path Reconstruction from Multi-Channel Audio of Echolocating Bats}},
year = {{2026}},
}