EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Two-Stage Wildlife Event Classification for Edge Deployment

Aditya S Viswanathan iD, Adis Bock, Zoe Bent, Mark A Peyton iD, Daniel M Tartakovsky iD, Javier E Santos

Sensors · 21.02.2026 · 26 (4) · 1366

DOI: 10.3390/s26041366 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 21.02.2026. Verfasst von Aditya S Viswanathan und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Energy Science and Engineering, Stanford University, Stanford, CA 94305, USA. Hauptthemen laut MEDLINE: Animals, Wild, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: United States Air Force Office of Scientific Research, Office of Advanced Scientific Computing Research, Air Force Office of Scientific Research.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Camera-based wildlife monitoring is often overwhelmed by non-target triggers and slowed by manual review or cloud-dependent inference, which can prevent timely intervention for high stakes human–wildlife conflicts. Our key contribution is a deployable, fully offline edge vision sensor that achieves near-real-time, highly accurate wildlife event classification by combining detector-based empty-image suppression with a lightweight classifier trained with a staged transfer-learning curriculum. Specifically, Stage 1 uses a pretrained You Only Look Once (YOLO)-family detector for permissive animal localization and empty-trigger suppression, and Stage 2 uses a lightweight EfficientNet-based binary classifier to confirm puma on detector crops and gate downstream actions. Our design is robust to low-quality nighttime monochrome imagery (motion blur, low contrast, illumination artifacts, and partial-body captures) and operates using commercially available components in connectivity-limited settings. In field deployments running since May 2025, end-to-end latency from camera trigger to action command is approximately 4 s. Ablation studies using a dataset of labeled wildlife images (pumas, not pumas) show that the two-stage approach substantially reduces false alarms in identifying pumas relative to a full-image classifier while maintaining high recall. On the held-out test set (N=1434 events), the proposed two-stage cascade achieves precision 0.983, recall 0.975, F1 0.979, accuracy 0.986, and balanced accuracy 0.983, with only 8 false positives and 12 false negatives. The system can be easily adapted for other species, as demonstrated by rapid retraining of the second stage to classify ringtails. Downstream responses (e.g., notifications and optional audio/light outputs) provide flexible actuation capabilities that can be configured to support intervention.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Animals, WildImage Processing, Computer-AssistedAnimalsClassification AlgorithmsDetection AlgorithmsAlgorithmsHumans

Stichwörter der Autor:innen

artificial intelligencecomputer visioncurriculum learningedge computingimage classificationobject detectionreal-time environmental monitoring

Autor:innen und Institutionen

  • Department of Energy Science and Engineering, Stanford University, Stanford, CA 94305, USA.
  • Pajarito Environmental Education Center, Los Alamos, NM 87544, USA.
  • Rubenstein School for Environment and Natural Resources, University of Vermont, Burlington, VT 05405, USA.
  • Department of Fish, Wildlife and Conservation Ecology, New Mexico State University, Las Cruces, NM 88003, USA.
  • Los Alamos National Laboratory, Earth and Environmental Sciences Division, Los Alamos, NM 87545, USA.

ORCID-Profile: Aditya S Viswanathan, Mark A Peyton, Daniel M Tartakovsky

Förderung

  • United States Air Force Office of Scientific Research
  • Office of Advanced Scientific Computing Research
  • Air Force Office of Scientific Research
  • Office of Advanced Scientific Computing Research within the Department of Energy Office of Science

Zitieren

APA

Viswanathan, A. S., Bock, A., Bent, Z., Peyton, M. A., Tartakovsky, D. M., & Santos, J. E. (2026). Two-Stage Wildlife Event Classification for Edge Deployment. Sensors, 26(4), 1366. https://doi.org/10.3390/s26041366

Vancouver

Viswanathan AS, Bock A, Bent Z, Peyton MA, Tartakovsky DM, Santos JE. Two-Stage Wildlife Event Classification for Edge Deployment. Sensors. 2026;26(4):1366. doi:10.3390/s26041366

BibTeX
@article{Viswanathan2026,
  author = {Viswanathan, Aditya S and Bock, Adis and Bent, Zoe and Peyton, Mark A and Tartakovsky, Daniel M and Santos, Javier E},
  title = {Two-Stage Wildlife Event Classification for Edge Deployment},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {4},
  pages = {1366},
  doi = {10.3390/s26041366}
}
Kostenlos starten →