About HALLO

Here we can include some information about the project generally. This section will not show up in the header menu above.

Humans and Algorithms Listening to Orcas explores the concept of training artificial intelligence systems to detect underwater whale vocalizations. The research goal is to develop a whale forecasting system to warn nearby ships of whale presence. This could prevent potentially fatal ship strikes for the endangered Southern Resident killer whales and other whale species that frequent the waters of the Salish Sea in the Pacific Northwest.

To train the artificial intelligence systems, we use recordings of verified whale vocalizations from several hydrophones located around the Salish Sea. These hydrophones are operated by various partners including citizen scientists and projects like the Orcasound project.

Our Research Progress

HALLO formed and began research efforts in 2019. Since then, we’ve been working on several objectives that work towards developing the real-time whale forecasting system in the Salish Sea.


Call Catalogue

We can introduce the catalogue project here. Give credit to John Ford & other contributors. Then probably link to its location.

We have compiled a catalogue of the 30 unique types of Southern Resident killer whale calls. This allows our team of annotators to manually identify individual call types contained in hydrophone recordings to train the artificial inteligence systems how to detect these calls.


Classification Models

We have models that provide classifications for heard sounds. For example, we have a binary classification model to determine whether a sound was an Killer Whale or not. These models were developed with data from Fisheries and Oceans Canada, JASCO, SMRU, … .

Check out the open source model’s on our public GitHub repository. There you can find instructions, scripts and configuration files required to train deep learning models at detecting and classifying vocalisations made by killer whales.


Forecast Models

We have more models! We bring together environmental data, citizen science observations, and hydrophone data to forecast where a cetacean is going to be. This system allows us to alert ports, ship captains, and … when a whale might be entering their path so they can take risk-reducing measures.

Killer whale by Photographer Name


Bioacoustics Workshop

In August 2022 we hosted a bioacoustics workshop (hybrid delivery).

Links to video recordings & slide presentations can be found here

The Team

This would be a good spot for a team photo – probably from a Zoom call.

Researchers

Dr. Ruth Joy, Principal Investigator

  • School of Environmental Science, Simon Fraser University

Dr. Steven Bergner

  • School of Computing Science, Simon Fraser University

Dr. Mike Dowd

  • Department of Mathematics and Statistics, Dalhousie University

Fabio Frazao

  • Department of Computer Science, Dalhousie University

Dr. Oliver Kirsebom

  • Open Ocean Robotics

Dr. Scott Viers

  • OrcaSound Collective, Beam Reach Marine Science and Sustainability School

Paul Nguyen Hong Duc

  • School of Mathematics and Statistics, Carleton University

Dr. Dave Campbell

  • School of Mathematics and Statistics, Carleton University

Collaborators

SMRU Consulting

  • Jason Wood

JASCO Applied Sciences

  • Jennifer Wladichuk, April Houweling, David Hannay

Ocean Network Canada

  • Jasper Kanes

Saturna Island Marine Research and Education Society (SIMRES)

  • Dr. Martin Wale

Digital Alliance Canada

  • Jillian Andersen

Logos of collaborators could probably be included here too.