Resumo: Passive monitoring of the soundscape (eco-acoustics) is a reliable and efficient tool for analyzing the state and composition of the landscape, which allows making decisions regarding the conservation and protection of biodiversity. In my thesis, I propose a methodology to analyze the soundscape of Colombian forests using unsupervised learning methods and establish a link with patterns related to connectivity. This proposal is based on the processing of audio recordings collected at multiple geographical points to obtain representative characteristics and patterns using spectral information and metadata from the recordings to later group the points, and thus establish a relationship between features, environmental variables, and patterns with connectivity. I hope the proposed methodology will make it possible to relate acoustic features of Colombian landscapes with patterns of connectivity modeled through heterogeneity as a measure of the structural and functional composition of the landscapes. Through my research, I aspire to contribute to the field of eco-acoustics by enhancing our understanding of acoustic patterns and their connection to ecological connectivity in Colombian landscapes.

Daniel Alexis Nieto Mora has a bachelor's degree in Mechatronic Engineering and a master's degree in Automation and Control from the Instituto Tecnológico Metropolitano (ITM) in Medellín, Colombia, and currently pursuing a Doctoral degree at ITM. His current interests are in machine learning algorithms, in particular in the characterization of acoustic data using autoencoders to effectively distinguish groups associated with different sound sources.

Os interessados em participar da palestra podem solicitar o link de transmissão através do email Este endereço de email está sendo protegido de spambots. Você precisa do JavaScript ativado para vê-lo.