Blog by the Media Technologies research group (GTM). Researching interactions between humans, machines and their environments.

08 July 2019 | Posted by Editorial Team GTM

GTM works in collaboration with UCM to classify Pipistrellus pipistrellus and Pipistrellus pygmaeus, two common species of bats in the Iberian Peninsula

GTM is working together with the Biodiversity, Ecology and Evolution Department from Universidad Complutense de Madrid (UCM) to classify automatically two common bat species in the Iberian Peninsula (classify Pipistrellus pipistrellus and Pipistrellus pygmaeus). Bats have an important role in the ecosystem, so determining their prevalence can contribute to their conservation. The study of bats through the analysis of their echolocation calls is a commonly used methodology. We have analyzed the calls of two bat species, Pipistrellus pipistrellus and Pipistrellus pygmaeus, because these two cryptic species are difficult to identify by their morphology features, but are more easily identified by their echolocation calls.

Figure 1: Pictures of Pipistrellus pipistrellus (left) and Pipistrellus pygmaeus (right). Images courtesy of Elena Tena (UCM).

Marta Bertran (GTM, winner of a Beca Colaboración 2018-19) and Elena Tena (PhD student from UCM), together with Dr. Alsina-Pagès (GTM), have analysed audio files obtained in real-life recordings by an Echo Meter Touch Pro 1 bat detector; time-expanded recordings of calls were first analyzed manually by means of their frequency, duration and inter-pulse interval. The first stage required the creation of a dataset with three classes (the two species and the silent intervals). Afterwards, two automatic bat detection and identification machine learning approaches are still under test, in a laboratory environment, which represents the previous step to real-life an urban scenario. This first stage of the collaboration has reached to its end by the defense of the Degree Thesis by Marta Bertran (GTM), but will keep working on the submission of a Journal Paper with the collaboration results.

Share

Add new comment

CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.
2 + 7 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.