On Enhancing Acoustic Event Detection by Using Feature Selection and Audiovisual Feature-Level Fusion

Acoustic Event Classification using spectral band selection and Non-Negative Matrix Factorization-based features

“…Good examples are the conventional Mel-Frequency Cepstral Coe cients (MFCC) (Temko & Nadeu, 2006;Zieger, 2008;Zhuang et al, 2010;Kwangyoun & Hanseok, 2011), log filter bank energies (Zhuang et al, 2010), Perceptual Linear Prediction (PLP) (Portelo et al, 2009), log-energy, spectral flux, entropy and zero-crossing rate (Temko & Nadeu, 2006;Perperis et al, 2011). The combination of some of these short-time features into high-dimensional acoustic vectors has also been studied, as well as the application of feature selection algorithms over these large pools of characteristics, in order to precisely reduce their dimensionality (Zhuang et al, 2008(Zhuang et al, , 2010Butko & Nadeu, 2010;Kiktova-Vozarikova et al, 2013).…”

Cough sound discrimination in noisy and reverberant environments using microphone arrays

“…In this noise estimator, mutually independent output signals were produced from the observed signals without any knowledge of the room acoustics or the direction of speech. The results of the noise estimator were compared to their previously developed noise estimator which used a null beamformer and a significant improvement was observed using the ICA-based noise estimator [33], Butko et al [34] attempted to detect various acoustic events that occurred in a meeting room in order to describe the various activities that might be occurring. The authors made use of microphone arrays as part of their research.…”

“…In the audio event detection (AED) system, a microphone array was used to provide spectro-temporal features along with audio source localization. The information extracted from the microphone arrays was further coupled with visual clues obtained from video cameras and fed into an HMM classifier in order to classify the appropriate acoustic event that had occurred [34],…”