Abstract

The automated acoustic detection of elephants is an important factor in alleviating the human-elephant conflict in Asia and Africa. In this paper, we present a method for the automated detection of elephant presence and evaluate it on a large dataset of wildlife recordings. We introduce a novel technique for signal enhancement to improve the robustness of the detector in noisy situations. Experiments show that the proposed detector outperforms existing methods and that signal enhancement strongly improves the robustness to noise sources from the environment. The proposed method is a first step towards an automated detection system for elephant presence.

Reference

Zeppelzauer, M., Stöger, A. S., & Breiteneder, C. (2013). Acoustic detection of elephant presence in noisy environments. In Proceedings of the 2nd ACM international workshop on Multimedia analysis for ecological data - MAED ’13. ACM, Austria. ACM. https://doi.org/10.1145/2509896.2509900