Automated ultrasonic vocalization analysis: Training and testing VocalMat on a rat-based dataset
Bibliographic record
Abstract
Background: Ultrasonic vocalizations (USVs) offer another way to study the behaviour of rodents in addition to commonly used visual methods. USV subtypes have been associated with behaviour such as the concurrence of 22-kHz calls and signs of distress (defensive behaviour). (1,2) However, the categories used to analyze USVs are a source of contention, most notably with 50-kHz calls, and may even be arbitrary altogether. (3) To facilitate subtyping calls, VocalMat has been developed for USV identification and classification, and it has shown an accuracy of greater than 98% for mice USV detection and 86% for mice USV classification. (4) In this project, we have constructed a rat-based dataset of USVs and then used it to train the VocalMat program to assess automated USV classification. Methods: Avisoft-SASLab Pro was used to manually classify USVs from 216 audio files. The sorted USVs were then used to train VocalMat’s classification program. Results: Our results show overall accuracies greater than 90% with the highest in the trill and flat categories (97.2% and 91.0%). We experimented with the number of USV categories and found high accuracies when grouping spectrographically similar calls, which are flat calls with up and down ramp calls (96.9%) and trill calls with trill jump and flat-trill calls (98.7%). Limitations: There are large variations in the number of calls per category in our dataset. More data is needed to fill these gaps and provide more training samples for infrequent calls. Conclusions: By creating a database of rat USVs and then using it to train VocalMat, we have shown the potential of its adaption to a rat vocal repertoire. Going forward, we hope to test more variations of USV categories on machine learning programs to establish a robust approach to classifying USVs.
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 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 | 0.003 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.007 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".