Open-source deep learning models for acoustic detection and classification of orcas
Bibliographic record
Abstract
Current efforts to acoustically study and monitor killer whales in British Columbia (BC), including the endangered population of Southern Resident killer whales (SRKW), are hampered by the lack of sufficiently accurate sound detection and classification algorithms. Recently, several research groups have reported significant improvements in algorithm performance utilizing deep neural networks. However, for most practitioners, these novel tools remain out of reach. To bridge this gap, we are developing a set of open-source deep learning models for acoustic detection and classification of BC's killer whales. These models will be made publicly available along with expert-curated training and test sets to facilitate further development and applications. We are collaborating with Orcasound to deploy the models on their live data. We are also working together with the PAMGuard developer team to ensure that our models can be seamlessly imported and used in PAMGuard, a widely used open-source software platform for passive acoustic monitoring. In this contribution, we will provide an overview of our deep learning methodology and present preliminary results on model performance, with particular attention to the model's ability to handle diverse and variable acoustic environments and generalize to new “unseen” environments.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.005 | 0.003 |
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".