Geology topics

Alexej P.K. Sirén

Publications and source records attributed to Alexej P.K. Sirén.

3 recordsLinked to original sources

USGS2026

Life-history stages and behavior influence demographic classification of moose captured on remote cameras

Obtaining accurate information on demographic states, such as the age and sex classes of animals, is an important step for monitoring wildlife populations. Traditionally, demographic data are collected from harvest, aerial surveys and telemetry studies. However, these methods can be expensive, limited to small spatial scales, or biased due to human behavior. Remote cameras have become a mainstay for studying and monitoring wildlife as they are relatively inexpensive, can be deployed over large spatial scales, and effort can be accounted for during surveys. For some species, a variety of demographic information, such as sex and age classes, can be obtained from pictures. Moose Alces alces are a photogenic species found across boreal and semi-boreal forests of the Northern Hemisphere. Previous studies have used demographic data from remote cameras to estimate demographic parameters and population dynamics. A primary assumption is that these age and sex classes are accurately classified. However, numerous factors can influence the ability of observers to identify age and sex classes of moose captured on cameras. We used data from 84 cameras from a 3-year period (2021–2024) in northern Maine, USA, to evaluate how temporal, environmental, site-level, and endogenous factors influence observers' ability to classify age and sex classes of moose. Using Bayesian categorical regression models, we found that temporal variability, position and proximity of moose from cameras, and the behavior of moose influenced our ability to identify age and sex classes. This information can be used to decide which periods to use data for population modeling and how to design studies to reduce the amount of uncertainty associated with different age and sex classes. We anticipate that our approach could also be used for other species whose age and sex classes can be differentiated using remote cameras.

Maine↗

USGS2025

DeepFaune New England: A species classification model for trail camera images in northeastern North America

The DeepFaune New England model classifies wildlife species in trail camera images, identifying 24 taxa from northeastern North America with high (97%) accuracy. The model was adapted from the DeepFaune model for identifying European wildlife, demonstrating the practicality of transfer learning across continents. The majority of training data is openly licensed, and the model itself is open source, enabling easy integration into camera trapping workflows. The open source software is available at ( https://code.usgs.gov/vtcfwru/deepfaune-new-england ), and has been further integrated into the PyTorch-Wildlife framework.

New England↗

USGS2023

Evaluating a tandem human-machine approach to labelling of wildlife in remote camera monitoring

Remote cameras (“trail cameras”) are a popular tool for non-invasive, continuous wildlife monitoring, and as they become more prevalent in wildlife research, machine learning (ML) is increasingly used to automate or accelerate the labor-intensive process of labelling (i.e., tagging) photos. Human-machine hybrid tagging approaches have been shown to greatly increase tagging efficiency (i.e., time to tag a single image). However, those potential increases hinge on the extent to which an ML model makes correct vs. incorrect predictions. We performed an experiment using a ML model that produces bounding boxes around animals, people, and vehicles in remote camera imagery (MegaDetector) to consider the impact of a ML model’s performance on its ability to accelerate human labeling. Six participants tagged trail camera images collected from 12 sites in Vermont and Maine, USA (January–September 2022) using three tagging methods (one with ML bounding box assistance and two without assistance). We used a generalized linear mixed model to examine the influence of ML model performance and tagging method on tagging efficiency. We found that ML bounding boxes offer significant improvement in tagging efficiency when labelling data compared to unassisted tagging. Additionally, the time taken to label with bounding boxes was not statistically different from an unassisted tagging approach. However, we found that gains in efficiency are contingent on the ML algorithm’s performance and that incorrect ML predictions, particularly the 4.2% false positive and 3.6% false negative predictions, can slow the tagging process compared to a non-hybrid approach. These findings indicate that although practitioners usually forgo the production of bounding boxes when selecting a data labelling process due to the increased effort, ML bounding box-assisted tagging can offer an efficient method for labeling. More broadly, ML-assisted data labelling offers an opportunity to accelerate the analysis of trail camera imagery, but an assessment of the ML model’s performance can illuminate whether the hybrid-tagging approach is ultimately a help or hinderance.

Maine, Vermont↗