Development of Automated Whistle and Click Classifiers for Odontocete Species in the Northwest Atlantic Ocean and the Waters Surrounding the Hawaiian Islands
Abstract
The primary objective of this effort is to develop combined automated whistle and click classifiers for odontocete species in the northwest Atlantic Ocean and the waters surrounding the Hawaiian Islands. We will also incorporate ancillary information (e.g. vocalization rate, relative abundance of whistles and clicks, latitude of acoustic encounter) about acoustic encounters into the classifiers in an effort to improve classification the performance. These classifiers will be implemented in an existing whistle classifier, known as the Real-time Odontocete Call Classification Algorithm (ROCCA). ROCCA currently is available to users as a module in the acoustic processing software platform, PAMGuard (Gillespie et al. 2008). We will integrate new classifiers with fully automated detectors and feature extractors in PAMGuard to provide a complete tool for efficiently and automatically processing the large datasets generated from Passive Acoustic Monitoring (PAM). When these tasks have been completed, we will use archival acoustic data recorded in the temperate Pacific Ocean to create whistle, click and ancillary information feature vectors. These feature vectors will be used to train a temperate Pacific classifier under an LMR-funded effort.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2014
- Accession Number
- ADA616536
Entities
People
- Julie N. Oswald
- Tina M. Yack