{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,24]],"date-time":"2025-12-24T10:10:27Z","timestamp":1766571027025,"version":"3.48.0"},"reference-count":40,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2022,5,19]],"date-time":"2022-05-19T00:00:00Z","timestamp":1652918400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Effective management of threatened and invasive species requires regular and reliable population estimates. Drones are increasingly utilised by ecologists for this purpose as they are relatively inexpensive. They enable larger areas to be surveyed than traditional methods for many species, particularly cryptic species such as koalas, with less disturbance. The development of robust and accurate methods for species detection is required to effectively use the large volumes of data generated by this survey method. The enhanced predictive and computational power of deep learning ensembles represents a considerable opportunity to the ecological community. In this study, we investigate the potential of deep learning ensembles built from multiple convolutional neural networks (CNNs) to detect koalas from low-altitude, drone-derived thermal data. The approach uses ensembles of detectors built from combinations of YOLOv5 and models from Detectron2. The ensembles achieved a strong balance between probability of detection and precision when tested on ground-truth data from radio-collared koalas. Our results also showed that greater diversity in ensemble composition can enhance overall performance. We found the main impediment to higher precision was false positives but expect these will continue to reduce as tools for geolocating detections are improved. The ability to construct ensembles of different sizes will allow for improved alignment between the algorithms used and the characteristics of different ecological problems. Ensembles are efficient and accurate and can be scaled to suit different settings, platforms and hardware availability, making them capable of adaption for novel applications.<\/jats:p>","DOI":"10.3390\/rs14102432","type":"journal-article","created":{"date-parts":[[2022,5,20]],"date-time":"2022-05-20T00:18:11Z","timestamp":1653005891000},"page":"2432","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":14,"title":["Automated Detection of Koalas with Deep Learning Ensembles"],"prefix":"10.3390","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6523-0304","authenticated-orcid":false,"given":"Megan","family":"Anschau","sequence":"first","affiliation":[{"name":"School of Biological and Environmental Science, Queensland University of Technology, 2 George Street, Brisbane, QLD 4000, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0983-5480","authenticated-orcid":false,"given":"Simon","family":"Denman","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering and Robotics, Queensland University of Technology, 2 George Street, Brisbane, QLD 4000, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0321-009X","authenticated-orcid":false,"given":"Evangeline","family":"Corcoran","sequence":"additional","affiliation":[{"name":"The Alan Turing Institute, 96 Euston Road, London NW1 2DB, UK"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8445-0575","authenticated-orcid":false,"given":"Grant","family":"Hamilton","sequence":"additional","affiliation":[{"name":"School of Biological and Environmental Science, Queensland University of Technology, 2 George Street, Brisbane, QLD 4000, Australia"}]}],"member":"1968","published-online":{"date-parts":[[2022,5,19]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"20191487","DOI":"10.1098\/rspb.2019.1487","article-title":"Optimizing future biodiversity sampling by citizen scientists","volume":"286","author":"Callaghan","year":"2019","journal-title":"Proc. R. Soc. B Biol. Sci."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"8176","DOI":"10.1002\/ece3.6522","article-title":"New technologies in the mix: Assessing N-mixture models for abundance estimation using automated detection data from drone surveys","volume":"10","author":"Corcoran","year":"2020","journal-title":"Ecol. Evol."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"586","DOI":"10.1071\/WR18034","article-title":"A comparison of unmanned aerial vehicles (drones) and manned helicopters for monitoring macropod populations","volume":"45","author":"Gentle","year":"2018","journal-title":"Wildl. Res."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"639","DOI":"10.1071\/WR18122","article-title":"Estimating kangaroo density by aerial survey: A comparison of thermal cameras with human observers","volume":"46","author":"Lethbridge","year":"2019","journal-title":"Wildl. Res."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"2623","DOI":"10.1080\/01431161.2017.1280639","article-title":"Adapting astronomical source detection software to help detect animals in thermal images obtained by unmanned aerial systems","volume":"38","author":"Longmore","year":"2017","journal-title":"Int. J. Remote Sens."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"23050","DOI":"10.1038\/s41598-021-02459-w","article-title":"Camera trap placement for evaluating species richness, abundance, and activity","volume":"11","author":"Tanwar","year":"2021","journal-title":"Sci. Rep."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"5504","DOI":"10.1080\/01431161.2017.1390621","article-title":"Exploring the feasibility of unmanned aerial vehicles and thermal imaging for ungulate surveys in forests\u2014Preliminary results","volume":"39","author":"Witczuk","year":"2018","journal-title":"Int. J. Remote Sens."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Hong, S.J., Han, Y., Kim, S.Y., Lee, A.Y., and Kim, G. (2019). Application of deep-learning methods to bird detection using unmanned aerial vehicle imagery. Sensors, 19.","DOI":"10.3390\/s19071651"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Leigh, C., Heron, G., Wilson, E., Gregory, T., Clifford, S., Holloway, J., McBain, M., Gonzalez, F., McGree, J., and Brown, R. (2019). Using virtual reality and thermal imagery to improve statistical modelling of vulnerable and protected species. PLoS ONE, 14.","DOI":"10.1101\/645291"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"101212","DOI":"10.1016\/j.ecoinf.2021.101212","article-title":"Advances in image acquisition and processing technologies transforming animal ecological studies","volume":"61","author":"Nazir","year":"2021","journal-title":"Ecol. Inform."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Prosekov, A., Kuznetsov, A., Rada, A., and Ivanova, S. (2020). Methods for monitoring large terrestrial animals in the wild. Forests, 11.","DOI":"10.3390\/f11080808"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"45127","DOI":"10.1038\/srep45127","article-title":"Automated detection and enumeration of marine wildlife using unmanned aircraft systems (UAS) and thermal imagery","volume":"7","author":"Seymour","year":"2017","journal-title":"Sci. Rep."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"434","DOI":"10.1002\/wsb.1090","article-title":"Evaluating the Use of Drones Equipped with Thermal Sensors as an Effective Method for Estimating Wildlife","volume":"44","author":"Beaver","year":"2020","journal-title":"Wildl. Soc. B"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"181","DOI":"10.1002\/wsb.629","article-title":"Visible and thermal infrared remote sensing for the detection of white-tailed deer using an unmanned aerial system","volume":"40","year":"2016","journal-title":"Wildl. Soc. B"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"898","DOI":"10.1111\/aje.12513","article-title":"Identification of African antelope species: Using thermographic videos to test the efficacy of real-time thermography","volume":"56","author":"Goodenough","year":"2018","journal-title":"Afr. J. Ecol."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"108598","DOI":"10.1016\/j.biocon.2020.108598","article-title":"When you can\u2019t see the koalas for the trees: Using drones and machine learning in complex environments","volume":"247","author":"Hamilton","year":"2020","journal-title":"Biol. Conserv."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"241","DOI":"10.5194\/isprsarchives-XL-1-W4-241-2015","article-title":"Wildlife multispecies remote sensing using visible and thermal infrared imagery acquired from an unmanned aerial vehicle (UAV)","volume":"XL-1\/W4","year":"2015","journal-title":"Int. Arch. Photogramm. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"3208","DOI":"10.1038\/s41598-019-39917-5","article-title":"Automated detection of koalas using low-level aerial surveillance and machine learning","volume":"9","author":"Corcoran","year":"2019","journal-title":"Sci. Rep."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1280","DOI":"10.1111\/2041-210X.12127","article-title":"Estimating multispecies abundance using automated detection systems: Ice-associated seals in the Bering Sea","volume":"5","author":"Conn","year":"2014","journal-title":"Methods Ecol. Evol."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"1103","DOI":"10.1111\/2041-210X.13581","article-title":"Automated detection of wildlife using drones: Synthesis, opportunities and constraints","volume":"12","author":"Corcoran","year":"2021","journal-title":"Methods Ecol. Evol."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"685","DOI":"10.1016\/j.tree.2015.08.008","article-title":"Emerging Technologies to Conserve Biodiversity","volume":"30","author":"Pimm","year":"2015","journal-title":"Trends Ecol. Evol."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"139","DOI":"10.1016\/j.rse.2018.06.028","article-title":"Detecting mammals in UAV images: Best practices to address a substantially imbalanced dataset with deep learning","volume":"216","author":"Kellenberger","year":"2018","journal-title":"Remote Sens. Environ."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"101547","DOI":"10.1016\/j.ecoinf.2021.101547","article-title":"Using Web images to train a deep neural network to detect sparsely distributed wildlife in large volumes of remotely sensed imagery: A case study of polar bears on sea ice","volume":"68","author":"Chabot","year":"2022","journal-title":"Ecol. Inform."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"445","DOI":"10.1002\/rse2.200","article-title":"21 000 birds in 4.5 h: Efficient large-scale seabird detection with machine learning","volume":"7","author":"Kellenberger","year":"2021","journal-title":"Remote Sens. Ecol. Conserv."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"369","DOI":"10.1002\/rse2.195","article-title":"Using very-high-resolution satellite imagery and deep learning to detect and count African elephants in heterogeneous landscapes","volume":"7","author":"Duporge","year":"2021","journal-title":"Remote Sens. Ecol. Conserv."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"407","DOI":"10.1080\/01431161.2018.1523580","article-title":"Hot monkey, cold reality: Surveying rainforest canopy mammals using drone-mounted thermal infrared sensors","volume":"40","author":"Kays","year":"2019","journal-title":"Int. J. Remote Sens."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Menikdiwela, M., Nguyen, C., Li, H., and Shaw, M. (2017, January 4\u20136). CNN-based small object detection and visualization with feature activation mapping. Proceedings of the 2017 International Conference on Image and Vision Computing New Zealand, Christchurch, New Zealand.","DOI":"10.1109\/IVCNZ.2017.8402455"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"104046","DOI":"10.1016\/j.imavis.2020.104046","article-title":"Deep learning-based object detection in low-altitude UAV datasets: A survey","volume":"104","author":"Mittal","year":"2020","journal-title":"Image Vision Comput."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"241","DOI":"10.1007\/s11704-019-8208-z","article-title":"A survey on ensemble learning","volume":"14","author":"Dong","year":"2020","journal-title":"Front. Comput. Sci."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"6690590","DOI":"10.1155\/2021\/6690590","article-title":"Accurate multilevel classification for wildlife images","volume":"2021","author":"Escalona","year":"2021","journal-title":"Comput. Intel. Neurosc."},{"key":"ref_31","first-page":"7769199","article-title":"An Ensemble of Fine-Tuned Convolutional Neural Networks for Medical Image Classification","volume":"21","author":"Kumar","year":"2017","journal-title":"IEEE J. Biomed. Health"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Morovati, M., Karami, P., and Amjas, F.B. (2020). Accessing habitat suitability and connectivity for the westernmost population of Asian black bear (Ursus thibetanus gedrosianus, Blanford, 1877) based on climate changes scenarios in Iran. PLoS ONE, 15.","DOI":"10.1371\/journal.pone.0242432"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"7591","DOI":"10.1002\/ece3.7591","article-title":"An automatic method for removing empty camera trap images using ensemble learning","volume":"11","author":"Yang","year":"2021","journal-title":"Ecol. Evol."},{"key":"ref_34","unstructured":"Ying, X. (2014). Ensemble Learning, University of Georgia. Available online: https:\/\/www.researchgate.net\/publication\/262369664_Ensemble_Learning."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"734990","DOI":"10.3389\/fdata.2021.734990","article-title":"Testing a generalizable machine learning workflow for aquatic invasive species on rainbow trout (Oncorhynchus mykiss) in Northwest Montana","volume":"4","author":"Carter","year":"2021","journal-title":"Front. Big Data"},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Giese, L., Melzheimer, J., Bockm\u00fchl, D., Wasiolka, B., Rast, W., Berger, A., and Wachter, B. (2021). Using machine learning for remote behaviour classification\u2014Verifying acceleration data to infer feeding events in free-ranging cheetahs. Sensors, 21.","DOI":"10.3390\/s21165426"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Yu, J., Su, H., Liu, J., Yang, Z., Zhang, Z., Zhu, Y., Yang, L., and Jiao, B. (2019, January 27\u201328). A strong baseline for tiger re-ID and its bag of tricks. Proceedings of the 2019 IEEE\/CVF International Conference on Computer Vision Workshop (ICCVW), Seoul, Korea.","DOI":"10.1109\/ICCVW.2019.00040"},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Padilla, R., Netto, S.L., and da Silva, E.A.B. (2020, January 1\u20133). A survey on performance metrics for object-detection algorithms. Proceedings of the 2020 International Conference on Systems, Signals and Image Processing, Rio de Janeiro, Brazil. Available online: https:\/\/ieeexplore.ieee.org\/document\/9145130.","DOI":"10.1109\/IWSSIP48289.2020.9145130"},{"key":"ref_39","unstructured":"Ferrari, V., Hebert, M., Sminchisescu, C., and Weiss, Y. (2018). Localization recall precision (LRP): A new performance metric for object detection. Computer Vision\u2014ECCV 2018, Lecture Notes in Computer Science, Springer."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"1893","DOI":"10.1007\/s11135-017-0574-8","article-title":"Saturation in qualitative research: Exploring its conceptualization and operationalization","volume":"52","author":"Saunders","year":"2018","journal-title":"Qual. Quant."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/14\/10\/2432\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,24]],"date-time":"2025-12-24T10:07:58Z","timestamp":1766570878000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/14\/10\/2432"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,5,19]]},"references-count":40,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2022,5]]}},"alternative-id":["rs14102432"],"URL":"https:\/\/doi.org\/10.3390\/rs14102432","relation":{},"ISSN":["2072-4292"],"issn-type":[{"type":"electronic","value":"2072-4292"}],"subject":[],"published":{"date-parts":[[2022,5,19]]}}}