A key part of the study of whales in southern places is finding and prompting experts to send ships to gather faecal samples or booty parts to better understand what whales eat.
Hydrophones, underwater microphones used to locate whales, are especially useful in the night or in bad weather when observation grids are ineffective. Computer algorithms play an increasing role in the analysis of hydrophone data, but human listeners can complement and improve these algorithms.
A research project known as Orcasound has produced a web application that will allow developing scientists to listen to the sound that is painted from the hydrophone near San Juan Island to identify whales and other new sounds.
Scott Veirs, a Seattle-based bioaccupist and lead researcher at the Orcasound project, will describe a new web application and the value of citizens' science at the 176th meeting of the American Acoustic Society, held in cooperation with the Canadian Acoustical Association 2018 Acoustical Week in Canada, 5-9 November at the Victoria Conference Center in Victoria, Canada.
Citizens' scientists were useful in discovering whales and spotting unusual activities, such as the presence of other animals or noise from transport traffic. Orcasound's goal is to provide a cheap and affordable way for people who are interested in researching and preserving marine life to participate in research, Veirs said. The issue at the heart of the project, he added, is how to organize and train people who listen to audio streaming to be better whale detectors. The Orcasound project also stores audio data on cloud storage servers for later analysis – both humans and algorithms.
Each node in the network uses a cheap Raspberri Pi computer with additional audio hardware. Computers run the Linux operating system and open source software for encoding and sending audio using standard data formats that are popular with video streaming services such as IouTube. This reduces costs while maximizing search engine compatibility and ease of use. "We want citizens to learn to listen to signals really easily," Veirs said.
Future versions of the app will contain buttons that users can click when they hear something interesting, which will help hide data for analysis algorithms later. Although there may be some friendly rivalry between machines and humans in this arena, the Orcasound application aims to bring synergy between citizen scientists and sophisticated algorithms.
How to reduce the impact of forest ships on fish? Slow them down
Presentation # 2pAO1, "Orcasound app: open source solution for streaming live ocean sound for citizen scientists and cloud-based algorithms", Scott Veirs will be held on Tuesday, November 6, at 1:00 am. in the Eskuimalt Victoria Conference Center in Victoria, British Columbia, Canada. More information on the project can be found at vvv.orcasound.net