Active Learning for Soundscape Classification

This repository contains Java source code for classifying long-duration environmental audio recordings using active learning. The active learning approaches are described and evaluated in:

Mahnoosh Kholghi, Yvonne Phillips, Michael Towsey, Laurianne Sitbon, and Paul Roe, “Active learning for classifying long-duration audio recordings of the environment”, Methods in Ecology and Evolution Journal, 2018.

This citation should be used in all publications that use data, concepts, or results generated from this code base:

Mahnoosh Kholghi. (2018). Active Learning for Soundscape Classification. http://doi.org/10.5281/zenodo.1288919