Title :
Hierarchical neural network classifier for an efficient incident detection based on sound content analysis
Author_Institution :
Bilgisayar Muhendisligi Bolumu, Mevlana Univ., Konya, Turkey
Abstract :
A sound content analysis is proposed to detect incident at intersections, which is suitable to implement on hardware such as FPGA. Due to confusion between the sound classes, an hierarchical classifier architecture is proposed to improve the classification performance. The proposed architecture and the feature extraction algorithm are suitable for parallel implementation.
Keywords :
acoustic signal detection; feature extraction; neural net architecture; signal classification; FPGA; classification performance; feature extraction algorithm; hierarchical neural network classifier architecture; incident detection; sound content analysis; Accidents; Field programmable gate arrays; Mel frequency cepstral coefficient; Neural networks; Safety; Transportation;
Conference_Titel :
Signal Processing and Communications Applications Conference (SIU), 2012 20th
Conference_Location :
Mugla
Print_ISBN :
978-1-4673-0055-1
Electronic_ISBN :
978-1-4673-0054-4
DOI :
10.1109/SIU.2012.6204697
