The University of Southampton
Engineering and the Environment

Research project: Methods for automatically classifying humpback whale calls

Currently Active: 
Yes

The songs produced by male humpback whales during the breeding season have been studied for the last four decades because their complexity suggests they play an important role in the population dynamics.

Project Overview

The songs produced by male humpback whales during the breeding season have been studied for the last four decades because their complexity suggests they play an important role in the population dynamics. This study aims at providing a tool to detect and classify humpback whale songs, which could be easily employed in real time whilst studying populations across the world. Indeed, humpback whales are a worldwide spread species whose populations present different songs repertoires which are constituted by the same building blocks. The latter can be associated in a variety of sequences which may vary from year to year and from one individual to another. The first objective achieved was to detect the building blocks of each song by developing an energy detector with a double threshold to identify the start and end of each sound unit. Manual classification was used to evaluate the performance of feature sets to determine which one was most suitable for the task. Results showed that Mel-frequency cepstrum coefficients characterise well the variety of humpback whale calls. At present, research is focused on the classification of the calls; in particular, supervised classification using Hidden Markov Models (HMMs), which should allow improving the classification. HMMs were used in a previous study to classify humpback whale calls with promising results but highlighting the need for an appropriate detection ahead of the classification task. By supervising the HMM clustering we expect to obtain an efficient tool to classify humpback whale calls efficiently, objectively and systematically.

Share this research project Share this on Facebook Share this on Google+ Share this on Twitter Share this on Weibo

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×