Bioacoustic Classification for Biodiversity Monitoring
Recognising birds, marine life and other animals in passive acoustic recordings to monitor biodiversity.
Tier Grades
Tier A: Frontier AI labs
feasible
Google builds the Perch and SurfPerch embedding models.
Tier B: Indonesian AI labs
feasible
Few-shot transfer from bird embeddings, as Perch shows, needs only small labelled sets of local species.
Tier C: Institution units such as ITB
feasible
SurfPerch shows a reef library at 2% of bird sample counts suffices, so a unit can build one for Indonesian reefs.
Tier D: Organisations such as SEMAI (PoT, PoC, PoV)
feasible
Released classifiers such as BirdNET process large survey audio without training, fitting environmental monitoring pilots.
Where this comes from
Tiers: a) frontier AI labs, b) Indonesian AI labs, c) institution units such as ITB, d) organisations such as SEMAI delivering proof of technology, concept and value. Each rationale cites the anchors below.
Anchors
- ComputeNot in the sources read: the Perch and SurfPerch abstracts give no training compute.
- DataPerch: bird vocalisation embeddings transfer to bat, marine mammal and amphibian calls; SurfPerch: ReefSet reef library at 2% of the sample count of bird libraries.
- Team size (proxy)Proxy only (author count of an exemplar paper, not a headcount): 4 authors on the Perch transfer study; 15 on SurfPerch.
- Time horizonMonths.
- Shelf-life signalsEmbeddings that classify most taxa few-shot, leaving only data collection
- Pipeline roleTier A releases embedding models; Tier C records and labels local species; Tier B scales them nationally; Tier D runs monitoring for plantations and conservation clients.
Where this comes from
Team size is a proxy (author counts of exemplar papers), not a headcount.
Who and What
Links
R&D TopicBioacoustic Classification for Biodiversity Monitoring
Models
Perch (Bird Vocalization Classifier)Google, audio classificationSurfPerchGoogle, audio classification
BirdNETBirdNET team, audio classificationHumpback whale detectorGoogle, audio classification
Perch (Bird Vocalization Classifier)Google, audio classificationSurfPerchGoogle, audio classificationBirdNETBirdNET team, audio classificationHumpback whale detectorGoogle, audio classificationUse Casesnone recorded
Who Runs Itnone recorded
Papers
- Global birdsong embeddings enable superior transfer learning for bioacoustic classificationBurooj Ghani et al., arXiv preprint, 2023 state of the art
- BEANS: The Benchmark of Animal SoundsMasato Hagiwara et al., arXiv preprint, 2022 state of the art
- AVES: Animal Vocalization Encoder based on Self-SupervisionMasato Hagiwara, arXiv preprint, 2022 state of the art
- Computational bioacoustics with deep learning: a review and roadmapDan Stowell, arXiv preprint, 2021 survey
Where this comes from
The papers to read first: where the direction began, its current best, and a survey when one exists.
Sources
- arxiv.org/abs/2307.06292 https://arxiv.org/abs/2307.06292
- arxiv.org/abs/2404.16436 https://arxiv.org/abs/2404.16436
- github.com/kahst/BirdNET-Analyzer https://github.com/kahst/BirdNET-Analyzer
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Bioacoustic Classification for Biodiversity Monitoring", https://aurion.semai.id/topics/bioacoustic-monitoring/, records of October 2026.