Computer-Assisted Photo-Identification of Narwhals
Bibliographic record
Abstract
Although the narwhal (Monodon monoceros) is economically and culturally important to northern residents, sound management of this species is impaired by large gaps in knowledge. Research on this species has been limited partly by the cost of the methods used, and partly because some of these methods are invasive and therefore condemned by Inuit communities. Photo-identification, a non-invasive, inexpensive, and easy-to-use method recently developed for narwhals, uses photographs of natural marks to identify individuals. Its main drawback is the extended time required to process photographs. We developed a computer program to accelerate the identification process and thus mitigate the main drawback of photo-identification. This program uses the locations of notches on the dorsal ridge to compare a new image to each individual in a catalogue and lists those individuals in decreasing order of similarity. We tested consistency in user assignment of dorsal ridge features and the accuracy of the program by comparing sets of known individuals. While assignment errors were common, the program ranked the true match within the first 10% of the catalogue 78% of the time. The program accelerates the matching process by 1.2 to 4.1 times for catalogues ranging in size from 40 to 500 individuals, and the degree of acceleration increases with the size of the catalogue. This program could also be applied to the beluga whale (Delphinapterus leucas), another important northern species.
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 | 0.001 | 0.001 |
| 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 | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.006 | 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".