Summary of published and open models related to acoustic signals from killer whales and other soniferous species heard within the Salish Sea (inland waters of southern BC, Canada, and inland waters of Washington State, USA) and across the range of Southern Resident Killer Whales (northern CA to southeast AK, USA). Resources related to model sharing are in the footer.

Southern Resident Killer Whale (SRKW) models

Binary call classifiers (any SRKW call or not)

• OrcaHello real time inference system (Overview | link to current model? | license)
• Pod.Cast annotation tool (Overview | link to current model? | license)
• OrcaAL = Orca Active Learning annotation tool (Overview | models in Orcasound's Acoustic Sandbox (S3 bucket) | 34 MB | license)

• Val's autoencoder + CNN ( Overview | Working repo | license)

Bigg's killer whale models

• None available as of Nov, 2022.

Humpback vocalization models

• Google/NOAA model (87 MB | Apache license)

Northern Resident Killer Whale (NRKW) models

• Deeke?
• Christian Bergler ORCA-SPOT & other Github repos

---

Model sharing resources

• Tensorflow common saved model API overview allows easy import of model into Python