Title :
Using hierarchical method to improve real time for audio-based surveillance system
Author_Institution :
Shandong Univ. of Finance & Econ., Jinan, China
Abstract :
The real-time is important to audio-based surveillance system. In this paper, we present a use of hierarchical method to improve the real-time for audio-based surveillance system. This method can improve the real-time by reducing the computation cost. Experiments conducted on different scenes of monitoring data show the benefits of proposed approach. The proposed method can improve the real-time from 0.1064xRT to 0.0197xRT in case of the performance degradation slightly.
Keywords :
audio signal processing; surveillance; abnormal sound classification; abnormal sound detection; audio-based surveillance system; hierarchical method; Acoustics; Computational modeling; Event detection; Hidden Markov models; Hierarchical systems; Real-time systems; Surveillance; Hierarchical method; abnormal sound detection and classification; audio-based surveillance;
Conference_Titel :
Chinese Spoken Language Processing (ISCSLP), 2014 9th International Symposium on
Conference_Location :
Singapore
DOI :
10.1109/ISCSLP.2014.6936607
