RT journal article
T1 A global assessment of BirdNET performance: Differences among continents, biomes, and species
A1 Funosas, David
A1 Sebastián-González, E.
A1 Morant, Jon
A1 Marín Gómez, O.H.
A1 Mendoza, Irene
A1 Mohedano Muñoz, Miguel A.
A1 Santamaría, E.
A1 Bastianelli, Giulia
A1 Marquez Rodríguez, Alba
A1 Budka, Michał
A1 Bota, G.
A1 Alonso-Moya, C.D.
A1 De la Peña Rubio, José M.
A1 de la Morena, E.L.G.
A1 Santa-Cruz, M.
A1 de la Nava, Pablo
A1 Fernández Tizón, Mario
A1 Sánchez Mateos, Hugo
A1 Barrero, Adrián
A1 Traba, Juan
A1 Osiejuk, Tomasz S.
A1 Hart, Patrick J
A1 Navine, Amanda K.
A1 Montoya Muñoz, Andrés F.
A1 de Araujo, Carlos B.
A1 Rosa, G.L.M.
A1 Torres, I.M.D.
A1 Catalano, Ana L.
A1 Simões, C.R.
A1 Llusia, Diego
A1 Morales, M.B.
A1 Acebes, Pablo
A1 Medina, J.A.
A1 Brown, N.
A1 Astaras, Christos
A1 Karmiris, I.
A1 Navarrete, E.
A1 Cauchoix, Maxime
A1 Barbaro, Luc
A1 Arend, D.
A1 Müeller, Sandra
A1 González-García, F.
A1 González Romero, Alberto
A1 Mammides, Christos
A1 Pontikis, Michaelangelo
A1 Jacuzzi, Giordano
A1 Olden, Julian D.
A1 Bombaci, Sara P.
A1 Marcacci, Gabriel
A1 Jacot, Alain
A1 Zurano, J.P.
A1 Gangenova, Elena
A1 Varela, D.
A1 di Sallo, Facundo
A1 Zurita, Gustavo A.
A1 Atemasov, Andrey
A1 Tremblay, Junior A.
A1 Lamarre, V.
A1 Hutschenreiter, A.
A1 Monroy Ojeda, Alan
A1 Díaz Vallejo, Mauricio
A1 Chaparro Herrera, Sergio
A1 Briers, Robert A.
A1 Sousa Lima, Renata
A1 Pinheiro, Thiago
A1 Da Silva, W.C.
A1 Calvente, Alice
A1 Paz, R.V.
A1 Salustio Gomes, Carlos
A1 Oliveira Junior, Dorgival D.
A1 Lima Santos, Cicero S.
A1 Pichorim, Mauro
A1 Molin, A.D.
A1 Antonelli, Alexandre
A1 Gogoleva, Svetlana
A1 Palko, Igo
A1 Trong, Hiếu V.
A1 Duarte, M.H.L.
A1 dos Santos Saturnino, Natalia
A1 Silva, S.R.
A1 Rainho, Ana
A1 Lopes, P.
A1 Schuchmann, Karl L.
A1 Marques, M.I.
A1 de Oliverira Tissiani, A.S.
A1 Littlewood, Nick A.
A1 Tuanmu, Mao-Ning
A1 Kepfer Rojas, Sebastian
A1 Aguilera, A.L.
A1 Brotons, Lluís
A1 Feldman, M.J.
A1 Imbeau, Louis
A1 Panwar, Pooja
A1 Weed, Aaron S.
A1 Dehwal, Anant
A1 Attisano, Alfredo
A1 Theuerkauf, Jörn
A1 Goodale, Eben
A1 Darras, K.F.A.
A1 Pérez Granados, Cristian
A2 Biología
K1 Confidence threshold
K1 Automated detection
K1 Passive acoustic monitoring
K1 Bird communities
K1 BirdNET
K1 Deep learning
AB Recent advances in machine learning have accelerated automated species detection across diverse ecological domains, enabling large-scale, non-invasive monitoring of biodiversity. In ornithological research, the combination of passive acoustic monitoring (PAM) and rapidly-developing novel identification tools such as BirdNET-a deep learning-based sound recognition algorithm-offers new opportunities for surveying vocally active bird communities. Here, we present the first worldwide evaluation of BirdNET using 4224 one-minute recordings from 67 sites across all continents annotated by local experts. More specifically, we assessed the capacity of BirdNET to accurately identify individual vocalizations and characterize bird communities based on the automated analysis of passively collected soundscapes. We further analyzed how its performance varies across continents, biomes, species, and minimum confidence thresholds. The proportion of correct BirdNET predictions (precision) was generally high and consistent across continents (range: 0.57-0.71) and biomes (range: 0.55-0.76). In contrast, the proportion of vocalizations successfully detected (recall) was generally lower and more heterogeneous across continents (range: 0.24-0.52) and biomes (range: 0.34-0.72), reflecting differences in species coverage and local ecological context. BirdNET predictive power, as measured by the Precision-Recall Area Under the Curve (PR AUC; higher values indicating better performance), was highest in North America, Oceania, and Europe (range: 0.16-0.23), moderate in Central/South America (0.13), and lowest in Africa and Asia (range: 0.03-0.04). Species-specific analyses revealed substantial heterogeneity in detection accuracy, with optimal confidence thresholds varying widely by species and analytical goal. Our results establish a global reference point for BirdNET reliability and highlight where algorithmic refinement and expanded acoustic sampling are most needed.
PB ELSEVIER
SN 1470-160X
YR 2026
FD 2026
LK http://hdl.handle.net/10498/40520
UL http://hdl.handle.net/10498/40520
LA eng
DS Repositorio Institucional de la Universidad de Cádiz
RD 08-nov-2026