Robust acoustic event classification using deep neural networks

Sharan, Roneel V.; Moir, T. J.

doi:10.1016/j.ins.2017.02.013

Cited by 40 publications

(22 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, an alternative two-dimensional time-frequency image method has been used [20], called the gammatonegram or cochleogram. While relatively new to the machine hearing field, this is derived from a well-established warping, the gammatone auditory filterbank function [17].…”

Section: Introductionmentioning

confidence: 99%

Time–Frequency Feature Fusion for Noise Robust Audio Event Classification

McLoughlin

Xie²,

Yan

et al. 2019

Circuits Syst Signal Process

View full text Add to dashboard Cite

This paper explores the use of three different two-dimensional time-frequency features for audio event classification with deep neural network back-end classifiers. The evaluations use spectrogram, cochleogram and constant-Q transform-based images for classification of 50 classes of audio events in varying levels of acoustic background noise, revealing interesting performance patterns with respect to noise level, feature image type and classifier. Evidence is obtained that two well-performing features, the spectrogram and cochleogram, make use of information that is potentially complementary in the input features. Feature fusion is thus explored for each pair of features, as well as for all tested features. Results indicate that a fusion of spectrogram and cochleogram information is particularly beneficial, yielding an impressive 50-class accuracy of over 96% in 0 dB SNR and exceeding 99% accuracy in 10 dB SNR and above. Meanwhile, the cochleogram image feature is found to perform well in extreme noise cases of − 5 dB and − 10 dB SNR.

show abstract

Section: Introductionmentioning

confidence: 99%

Time–Frequency Feature Fusion for Noise Robust Audio Event Classification

McLoughlin

Xie²,

Yan

et al. 2019

Circuits Syst Signal Process

View full text Add to dashboard Cite

show abstract

“…Numerous audio features and classifiers have been proposed for use in animal vocalization detection and classification [3]. In particular, Mel frequency cepstrum coefficients (MFCCs) with the classifier such as support vector machine (SVM), hidden Markov model (HMM), or deep neural network (DNN) were commonly used features in animal sound recognition [24][25][26][27] because of its strong ability in sound distinction and robustness. However, MFCCs only reflected the static characteristics of acoustic signals.…”

Section: Introductionmentioning

confidence: 99%

A Novel Method for Broiler Abnormal Sound Detection Using WMFCC and HMM

Liu

Zhao

et al. 2020

Journal of Sensors

View full text Add to dashboard Cite

Broilers produce abnormal sounds such as cough and snore when they suffer from respiratory diseases. The aim of this research work was to develop a method for broiler abnormal sound detection. The sounds were recorded in a broiler house for one week (24/7). There were 20 thousand white feather broilers reared on the floor in a building. Results showed that the developed recognition algorithm, using wavelet transform Mel frequency cepstrum coefficients (WMFCCs), correlation distance Fisher criterion (CDF), and hidden Markov model (HMM), provided an average accuracy, precision, recall, and F1 of 93.8%, 94.4%, 94.1%, and 94.2%, respectively, for broiler sound samples. The results indicate that sound analysis can be used in broiler respiratory assessment in a commercial broiler farm.

show abstract

“…ese improved MFCC features are evaluated on DNN and I-Vector classifiers where the overall accuracies are 58.98% and 56.13%, respectively. Sharan and Moir [7] compared two classification techniques (DNN and SVM) using single and combined feature vectors for robust sound classifications.…”

Section: Introductionmentioning

confidence: 99%

DGR: Gender Recognition of Human Speech Using One-Dimensional Conventional Neural Network

Alkhawaldeh

2019

Scientific Programming

View full text Add to dashboard Cite

The speech entailed in human voice comprises essentially paralinguistic information used in many voice-recognition applications. Gender voice is considered one of the pivotal parts to be detected from a given voice, a task that involves certain complications. In order to distinguish gender from a voice signal, a set of techniques have been employed to determine relevant features to be utilized for building a model from a training set. This model is useful for determining the gender (i.e., male or female) from a voice signal. The contributions are three-fold including (i) providing analysis information about well-known voice signal features using a prominent dataset, (ii) studying various machine learning models of different theoretical families to classify the voice gender, and (iii) using three prominent feature selection algorithms to find promisingly optimal features for improving classification models. The experimental results show the importance of subfeatures over others, which are vital for enhancing the efficiency of classification models’ performance. Experimentation reveals that the best recall value is equal to 99.97%; the best recall value is 99.7% for two models of deep learning (DL) and support vector machine (SVM), and with feature selection, the best recall value is 100% for SVM techniques.

show abstract

Robust acoustic event classification using deep neural networks

Cited by 40 publications

References 23 publications

Time–Frequency Feature Fusion for Noise Robust Audio Event Classification

Time–Frequency Feature Fusion for Noise Robust Audio Event Classification

A Novel Method for Broiler Abnormal Sound Detection Using WMFCC and HMM

DGR: Gender Recognition of Human Speech Using One-Dimensional Conventional Neural Network

Contact Info

Product

Resources

About