Application of Machine Learning for Automating Behavioral Tracking of Captive Bornean Orangutans ( Pongo Pygmaeus ).

Autor: Gammelgård F; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark., Nielsen J; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark., Nielsen EJ; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark., Hansen MG; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark., Alstrup AKO; Department of Nuclear Medicine & PET, Aarhus University Hospital and Department of Clinical Medicine, Aarhus University, Palle Juul Jensens Boulevard 99, 8000 Aarhus, Denmark., Perea-García JO; Faculty of Social and Behavioural Sciences, Leiden University, 2333 Leiden, The Netherlands., Jensen TH; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark.; Aalborg Zoo, Mølleparkvej 63, 9000 Aalborg, Denmark., Pertoldi C; Department of Chemistry and Bioscience, Aalborg University, Frederik Bajers Vej 7H, 9220 Aalborg, Denmark.; Aalborg Zoo, Mølleparkvej 63, 9000 Aalborg, Denmark.
Jazyk: angličtina
Zdroj: Animals : an open access journal from MDPI [Animals (Basel)] 2024 Jun 08; Vol. 14 (12). Date of Electronic Publication: 2024 Jun 08.
DOI: 10.3390/ani14121729
Abstrakt: This article applies object detection to CCTV video material to investigate the potential of using machine learning to automate behavior tracking. This study includes video tapings of two captive Bornean orangutans and their behavior. From a 2 min training video containing the selected behaviors, 334 images were extracted and labeled using Rectlabel. The labeled training material was used to construct an object detection model using Create ML. The use of object detection was shown to have potential for automating tracking, especially of locomotion, whilst filtering out false positives. Potential improvements regarding this tool are addressed, and future implementation should take these into consideration. These improvements include using adequately diverse training material and limiting iterations to avoid overfitting the model.
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje