Likelihood‐based photograph identification: Application with photographs of free‐ranging bison
Bibliographic record
Abstract
ABSTRACT Using photographs to identify individual animals and monitor populations is becoming more common. However, photographic identification methods where measurements of morphological traits (e.g., horn length) are compared have received little attention. We present an approach for aiding with the identification of individual animals from photographs. The approach incorporates measurement data, metadata from photographs, and multiple sources of error, and calculates a matching score between pairs of photographs using a likelihood‐based algorithm. We tested and identified the false‐rejection error rate using 91 photographs, representing 33 known free‐ranging bison ( Bison bison ), and 117 simulated data sets with varying numbers of unique individuals, morphological measurements, and photograph error. We then used the approach to estimate the adult population size of bison in Prince Albert National Park, Canada, in 2011. For bison, the false‐rejection rate of our approach was 0.055. Using a Huggins closed population model with misidentification, we estimated 103 (95% CI = 82–130) and 46 (95% CI = 37–58) adult female and male bison, respectively. After incorporating field‐based calf‐ and juvenile‐to‐female ratios, we estimated 202 (95% CI = 171.6–231.4) bison. We found this estimate to be plausible using 2 minimum‐count aerial surveys conducted in March 2011 and 2012 for comparison. With our approach, researchers and managers can build capture histories of individuals, which can be used for studies of population dynamics and habitat selection. This approach can incorporate any morphological measurements extracted from photographs (e.g., coat color), making it robust to a variety of species and study systems. © 2013 The Wildlife Society.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.011 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".