Impact of deep learning and post-processing algorithms performances on biodiversity metrics assessed on videos

Guardado en:
書目詳細資料
發表在:PLoS One vol. 20, no. 8 (Aug 2025), p. e0327577
主要作者: Valentine Fleuré
其他作者: Planolles, Kévin, Claverie, Thomas, Mulot, Baptiste, Villéger, Sébastien
出版:
Public Library of Science
主題:
在線閱讀:Citation/Abstract
Full Text
Full Text - PDF
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!

MARC

LEADER 00000nab a2200000uu 4500
001 3238652879
003 UK-CbPIL
022 |a 1932-6203 
024 7 |a 10.1371/journal.pone.0327577  |2 doi 
035 |a 3238652879 
045 2 |b d20250801  |b d20250831 
084 |a 174835  |2 nlm 
100 1 |a Valentine Fleuré 
245 1 |a Impact of deep learning and post-processing algorithms performances on biodiversity metrics assessed on videos 
260 |b Public Library of Science  |c Aug 2025 
513 |a Journal Article 
520 3 |a Assessing the escalating biodiversity crisis, driven by climate change, habitat destruction, and exploitation, necessitates efficient monitoring strategies to assess species presence and abundance across diverse habitats. Video-based surveys using remote cameras are a promising, non-invasive way to collect valuable data in various environments. Yet, the analysis of recorded videos remains challenging due to time and expertise constraints. Recent advances in deep learning models have enhanced image processing capabilities in both object detection and classification. However, the impacts on models’ performances and usage on assessment of biodiversity metrics on videos is yet to be assessed. This study evaluates the impacts of video processing rates, detection and identification model performance, and post-processing algorithms on the accuracy of biodiversity metrics, using simulated remote videos of fish communities and 14,406 simulated automated processing pipelines. We found that a processing rate of one image per second minimizes errors while ensuring detection of all species. However, even near-perfect detection (both recall and precision of 0.99) and identification (accuracy of 0.99) models resulted in overestimation of total abundance, species richness and species diversity due to false positives. We reveal that post-processing model outputs using a confidence threshold approach (i.e., to discard most erroneous predictions while also discarding a smaller proportion of correct predictions) is the most efficient method to accurately estimate biodiversity from videos. 
653 |a Climate change 
653 |a Deep learning 
653 |a Algorithms 
653 |a Habitat loss 
653 |a Pattern recognition 
653 |a Biodiversity 
653 |a Species diversity 
653 |a Image processing 
653 |a Automation 
653 |a Machine learning 
653 |a Accuracy 
653 |a Cameras 
653 |a Video 
653 |a Environmental degradation 
653 |a Species richness 
653 |a Classification 
653 |a Binomial distribution 
653 |a Information processing 
653 |a Object recognition 
653 |a Environmental 
700 1 |a Planolles, Kévin 
700 1 |a Claverie, Thomas 
700 1 |a Mulot, Baptiste 
700 1 |a Villéger, Sébastien 
773 0 |t PLoS One  |g vol. 20, no. 8 (Aug 2025), p. e0327577 
786 0 |d ProQuest  |t Health & Medical Collection 
856 4 1 |3 Citation/Abstract  |u https://www.proquest.com/docview/3238652879/abstract/embedded/6A8EOT78XXH2IG52?source=fedsrch 
856 4 0 |3 Full Text  |u https://www.proquest.com/docview/3238652879/fulltext/embedded/6A8EOT78XXH2IG52?source=fedsrch 
856 4 0 |3 Full Text - PDF  |u https://www.proquest.com/docview/3238652879/fulltextPDF/embedded/6A8EOT78XXH2IG52?source=fedsrch