A comparison of methods for detecting right whale calls
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
North Atlantic, North Pacific, and southern right whales all produce the up call, a frequency-modulated upsweep in the 50-200 Hz range.This call is one o f the most common sounds, and frequently the most common sound, received from right whales, and as such is a useful indicator o f the presence o f right whales for acoustic surveys.A data set was prepared o f 1857 calls and 6359 non-call sounds recorded from North Atlantic right whales (Eubalaena glacialis) near Georgia and Massachusetts.Two methods for the detection o f the calls were compared: spectrogram correlation and a neural network.Spectrogram correlation parameters were chosen two ways, by manual choice using a sample o f 20 calls, and by an optimization procedure that used all available calls.Neural network weights were trained via backpropagation on 9/10 o f the test data set.Performance was measured separately for calls of different signal-to-noise ratio, as SNR heavily influences the performance o f any detector.Results showed that the neural network performed best at this task, achieving an error rate o f less than 6%, and is thus the preferred detection method here.Spectrogram correlation may be useful in situations in which a large set o f training data is not available, as manual training on a small set o f examples achieved an error rate (26%) that may be acceptable for many applications. s o m m a i r eLes baleines franches de l'Atlantique Nord, du Pacific Nord et Sud produisent toutes une vocalisation montante, soit un balayage ascendant modul en frquence dans la rgion de 50 200 Hz.Cette vocalisation est un des sons les plus communs produit par les baleines franches et, par le fait mme, est un indicateur trs utile de la prsence des baleines lors de sondages acoustiques.Un ensemble de donnes a t prpar avec 1857 vocalisations et 6359 sons non vocaliss enregistrs auprs de baleines franches de l 'Atlantique Nord (Eubalaena glacialis) prs de la Georgie et du Massachusetts.Deux mthodes de dtection des vocalisations ont t compares: la corrlation de spectrogramme et le rseau neuronal.Les paramtres de la corrlation de spectrogramme ont t choisis de deux faons: par choix manuel, en utilisant seulement 20 vocalisations, et par une optimisation de la procdure utilisant toutes les vocalisations.Les coefficients de pondration du rseau neuronal ont t tabli par rtropropagation sur 9/10 des donnes de test.Les performances ont t mesures sparment pour des vocalisations ayant des rapports signal sur bruit diffrents, le rapport signal sur bruit ayant une grande influence sur tout dtecteur.Les rsultats dmontrent que le rseau neuronal performe mieux dans ce genre de tche, atteignant un taux d 'erreur de moins de 6% et, par consquent, est dfini ici comme la meilleure mthode de dtection.La corrlation de spectrogramme peut tre utile dans les situations o un grand nombre de donnes de formation ne sont pas disponibles.Le choix manuel sur de petite tranche d 'chantillons a atteint un taux d 'erreur (26%) qui pourrait tre acceptable dans plusieurs applications.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it