Automatic recognition of sound events can be valuable for efficient analysis of audio scenes. For example, detecting human activities like trespassing and hunting in natural environments can play an important role in their preservation by alerting authorities to take action. In the proposed system, each sound class is represented by a hidden Markov model created from descriptors in the time, frequency, and wavelet domains. The system has the ability to automatically adapt to acoustic conditions of different scenes via the feedback loop that refines an unsupervised model. A reliable testing process was adopted for assessing the performance of the system under adverse conditions characterized by highly nonstationary environmental noise.

Authors: Ntalampiras, Stavros; Potamitis, Ilyas; Fakotakis, Nikos
Affiliations: Politecnico di Milano, Milan, Italy; Technological Educational Institute of Crete, Rethymno, Greece; University of Patras, Patras, Greece(See document for exact affiliation information.)
JAES Volume 60 Issue 9 pp. 686-695; September 2012
Publication Date: October 9, 2012 Import into BibTeX
Permalink: https://www.aes.org/e-lib/browse.cfm?elib=16373

Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $33 for non-members and is free for AES members and E-Library subscribers.

Learn more about the AES E-Library

E-Library Location: (CD JAES60)   /jaes60/9/pg686.pdf

Start a discussion about this paper!