Passive Acoustic Monitoring for the Detection and Identification of Marine Mammals
Abstract
This project is intended to advance the state of passive acoustic monitoring of marine mammals. Improved methods of identifying cetaceans are developed to contribute to the Navy's mitigation efforts. The project is a multi-pronged study to advance the state of the field in three areas. The development of automated auditory scene analysis for delphinid tonal calls will permit subsequent work by investigators to exploit the use of whistles for classification and localization. Our approach is to dynamically build hypothesis graphs to represent overlapping whistles. The whistle graphs are disambiguated using information from both sides of the crossings. In parallel to this effort, two modeling techniques are being pursued to improve existing passive acoustic monitoring capabilities based on echolocation clicks of odontocetes. The first of these examines the use of a universal background model as proposed by Reynolds et al. (2000) for human speaker verification tasks. Reynolds' problem is how can one reject vocalizations for which there are no data to create a model. We adapt his concept of a universal background model by training a generalized odontocete model using data from a number of species. Using Bayesian learning, training data from a specific species adapts the parameters of the generalized model. By having elements of the generalized model in both the adapted and background models, elements from an unknown species are more likely to be a better match for the general model, and those from the targeted species that differ from the training data are less penalized. The second approach for echolocation clicks exploits recent machine learning work on submanifold learning. To detect and classify odontocetes, features or poignant characteristics of their signals must be extracted from the audio signal. Classification techniques must attempt to infer information about the producer of the signal (e.g., species) through a typically higher order set of features.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2011
- Accession Number
- ADA598456
Entities
People
- Marie A. Roch
Organizations
- San Diego State University