Random forest classification based acoustic event detection utilizing contextual-information and bottleneck features. (September 2018)

• Record Type: Journal Article
• Authors: Xia, Xianjun; Togneri, Roberto; Sohel, Ferdous; Huang, David
• Is Part Of: Pattern recognition. Volume 81(2018:Sep.)
• Page Start: 1
• Page End: 13
• Publication Date: 2018-09
• Subjects: Acoustic event detection -- Contextual information -- Global bottleneck features -- Category-specific bottleneck features; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2018.03.025
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A random forest classification based acoustic event detection system was constructed as the baseline system. Contextual information was employed to cope with the acoustic signals with long duration. Global bottleneck features were employed in the acoustic event detection system to utilize the prior knowledge of the event category information. Category-specific bottleneck features were employed in the acoustic event detection system to utilize the prior knowledge of the event boundary information. Evaluations on the UPC-TALP and ITC-IRST databases of highly. Abstract: The variety of event categories and event boundary information have resulted in limited success for acoustic event detection systems. To deal with this, we propose to utilize the long contextual information, low-dimensional discriminant global bottleneck features and category-specific bottleneck features. By concatenating several adjacent frames together, the use of contextual information makes it easier to cope with acoustic signals with long duration. Global and category-specific bottleneck features can extract the prior knowledge of the event category and boundary, which is ideally matched by the task of an event detection system. Evaluations on the UPC-TALP and ITC-IRST databases of highly variable acoustic events demonstrate the effectiveness of the proposed approaches by achieving a 5.30% and 4.44% absolute error rate improvement respectively compared to the state of art technique.