Testing the Effectiveness of Automated Acoustic Sensors for Monitoring Vocal Activity of Marbled Murrelets Brachyramphus Marmoratus
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
Cryptic nest sites and secretive breeding behavior make population estimates and monitoring of Marbled Murrelets Brachyramphus marmoratus difficult and expensive.Standard audio-visual and radar protocols have been refined but require intensive field time by trained personnel.We examined the detection range of automated sound recorders (Song Meters; Wildlife Acoustics Inc.) and the reliability of automated recognition models ("recognizers") for identifying and quantifying Marbled Murrelet vocalizations during the 2011 and 2012 breeding seasons at Kodiak Island, Alaska.The detection range of murrelet calls by Song Meters was estimated to be 60 m.Recognizers detected 20 632 murrelet calls (keer and keheer) from a sample of 268 h of recordings, yielding 5 870 call series, which compared favorably with human scanning of spectrograms (on average detecting 95% of the number of call series identified by a human observer, but not necessarily the same call series).The false-negative rate (percentage of murrelet call series that the recognizers failed to detect) was 32%, mainly involving weak calls and short call series.False-positives (other sounds included by recognizers as murrelet calls) were primarily due to complex songs of other bird species, wind and rain.False-positives were lower in forest nesting habitat (48%) and highest in shrubby vegetation where calls of other birds were common (97%-99%).Acoustic recorders tracked spatial and seasonal trends in vocal activity, with higher call detections in high-quality forested habitat and during late July/early August.Automated acoustic monitoring of Marbled Murrelet calls could provide cost-effective, valuable information for assessing habitat use and temporal and spatial trends in nesting activity; reliability is dependent on careful placement of sensors to minimize false-positives and on prudent application of digital recognizers with visual checking of spectrograms.
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 | 0.001 | 0.002 |
| 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 | 0.001 |
| 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