SpectroBird: Event-Level Vocalization Detection and Curation for Avian Ecology
- 1School of Biological Sciences, Georgia Institute of Technology, Atlanta, United States
Abstract
Passive acoustic monitoring often summarizes long recordings as fixed-window presence predictions, but many ecological questions require records of individual vocal events: how many occurred, when they occurred, whether they overlapped, how long they lasted, what frequency band they occupied, and whether each event was complete enough for structural measurement. We address this event-level measurement problem with SpectroBird, a spectrogram-based method for localizing and curating avian vocalizations in noisy field recordings. SpectroBird first applies a detection-localization model to identify candidate vocalizations as time-frequency boxes. An event-consolidation step merges repeated detections from overlapping windows while preserving overlapping vocalizations as separate events. A completeness-curation model then classifies each consolidated event as complete or incomplete before downstream event-level analysis. We demonstrate SpectroBird on breeding-season Hume's Leaf Warbler buzz vocalizations recorded in the Himalaya across a 3,200-3,800 m elevational gradient. Under shared window-level scoring, the detection-localization model remained competitive with AST and YOLOv11-cls baselines while retaining event timing and frequency bounds; event-level evaluation achieved AP > 0.999 for complete buzzes. Applied to two field seasons, the resulting event records showed that duration of complete buzzes declined with elevation, complete-buzz occurrence showed no simple elevation trend, and relative day-of-year occurrence patterns were strongly correlated between years. These results show how event-level detection and curation can support ecological analyses that depend on individual vocalizations rather than fixed-window presence alone.
Related
- A Vision-Language Model as a Teacher for Bird Vocalization Detection
- Species ecology, behavior, and substrate type shape eDNA detection of mammals at savanna waterholes
- Ghrelin inhibits courtship ultrasonic vocalization in male mice
- NED-Net: an annotation-training loop for rodent EEG event detection
- Aging in Autism: An Event-Related Potentials Study of Auditory Change Detection.
BibTeX
@misc{dreyer2026spectrobird,
title = {{SpectroBird: Event-Level Vocalization Detection and Curation for Avian Ecology}},
author = {E. G. Dreyer and P. Gokhale and V. Chandran and T. Fox and A. Hushcha and V. B. Prema and C. R. Clark and Y. Iravantchi and M. Postiglione and B. Shi and S. Mussmann and B. Freeman},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.13.758621},
url = {https://www.biorxiv.org/content/10.64898/2026.10.13.758621v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - SpectroBird: Event-Level Vocalization Detection and Curation for Avian Ecology
AU - E. G. Dreyer
AU - P. Gokhale
AU - V. Chandran
AU - T. Fox
AU - A. Hushcha
AU - V. B. Prema
AU - C. R. Clark
AU - Y. Iravantchi
AU - M. Postiglione
AU - B. Shi
AU - S. Mussmann
AU - B. Freeman
PY - 2026
DA - 2026/10/13
PB - bioRxiv
DO - 10.64898/2026.10.13.758621
UR - https://www.biorxiv.org/content/10.64898/2026.10.13.758621v1
AB - Passive acoustic monitoring often summarizes long recordings as fixed-window presence predictions, but many ecological questions require records of individual vocal events: how many occurred, when they occurred, whether they overlapped, how long they lasted, what frequency band they occupied, and whether each event was complete enough for structural measurement. We address this event-level measurement problem with SpectroBird, a spectrogram-based method for localizing and curating avian vocalizations in noisy field recordings. SpectroBird first applies a detection-localization model to identify candidate vocalizations as time-frequency boxes. An event-consolidation step merges repeated detections from overlapping windows while preserving overlapping vocalizations as separate events. A completeness-curation model then classifies each consolidated event as complete or incomplete before downstream event-level analysis. We demonstrate SpectroBird on breeding-season Hume's Leaf Warbler buzz vocalizations recorded in the Himalaya across a 3,200-3,800 m elevational gradient. Under shared window-level scoring, the detection-localization model remained competitive with AST and YOLOv11-cls baselines while retaining event timing and frequency bounds; event-level evaluation achieved AP > 0.999 for complete buzzes. Applied to two field seasons, the resulting event records showed that duration of complete buzzes declined with elevation, complete-buzz occurrence showed no simple elevation trend, and relative day-of-year occurrence patterns were strongly correlated between years. These results show how event-level detection and curation can support ecological analyses that depend on individual vocalizations rather than fixed-window presence alone.
ER -