Noise robust sound event classification with convolutional neural network

Özer, İlyas; Ozer, Zeynep; Fındık, Oğuz

doi:10.1016/j.neucom.2017.07.021

Cited by 85 publications

(39 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Next, the grayscale image is quantized into its RGB components; the mapping type is the

in MATLAB r2016b [ 38 ]. The mapping is expressed as

where

is the RGB spectrogram image and

is the non-linear

quantization function [ 32 ]. It is worth noting that, to facilitate the observation and analysis of the RGB spectrogram image, we deploy the color mapping in this paper.…”

Section: Signal Model and Time–frequency Analysis Methodsmentioning

confidence: 99%

“…First, the time–frequency analysis method based on the windowed short-time Fourier transform (STFT) [ 31 ] is employed to generate the spectrum of the MIMO-modulated signals. Then, the spectrum with different time windows is converted to a grayscale image, and this grayscale image is further transferred to the RGB spectrogram image [ 32 ]. Second, a fine-tuned AlexNet-based convolutional neural network (CNN) model is introduced to learn the features from the RGB spectrogram images.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Time–Frequency-Analysis-Based Blind Modulation Classification for Multiple-Antenna Systems

Jiang

Wang

et al. 2021

Sensors

View full text Add to dashboard Cite

Blind modulation classification is an important step in implementing cognitive radio networks. The multiple-input multiple-output (MIMO) technique is widely used in military and civil communication systems. Due to the lack of prior information about channel parameters and the overlapping of signals in MIMO systems, the traditional likelihood-based and feature-based approaches cannot be applied in these scenarios directly. Hence, in this paper, to resolve the problem of blind modulation classification in MIMO systems, the time–frequency analysis method based on the windowed short-time Fourier transform was used to analyze the time–frequency characteristics of time-domain modulated signals. Then, the extracted time–frequency characteristics are converted into red–green–blue (RGB) spectrogram images, and the convolutional neural network based on transfer learning was applied to classify the modulation types according to the RGB spectrogram images. Finally, a decision fusion module was used to fuse the classification results of all the receiving antennas. Through simulations, we analyzed the classification performance at different signal-to-noise ratios (SNRs); the results indicate that, for the single-input single-output (SISO) network, our proposed scheme can achieve 92.37% and 99.12% average classification accuracy at SNRs of −4 and 10 dB, respectively. For the MIMO network, our scheme achieves 80.42% and 87.92% average classification accuracy at −4 and 10 dB, respectively. The proposed method greatly improves the accuracy of modulation classification in MIMO networks.

show abstract

“…Next, the grayscale image is quantized into its RGB components; the mapping type is the

in MATLAB r2016b [ 38 ]. The mapping is expressed as

where

is the RGB spectrogram image and

is the non-linear

quantization function [ 32 ]. It is worth noting that, to facilitate the observation and analysis of the RGB spectrogram image, we deploy the color mapping in this paper.…”

Section: Signal Model and Time–frequency Analysis Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Time–Frequency-Analysis-Based Blind Modulation Classification for Multiple-Antenna Systems

Jiang

Wang

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…CNN is a typical multi-layer neural network first proposed for computer vision problems and it is very suitable for image-related applications in machine learning. In recent years, the application of CNN to environment sound classification has been widely reported [22][23][24][25]. We compute STFT spectrogram in each sound frame and then feed it as an image input to the CNN classifier.…”

Section: Classical Machine Learningmentioning

confidence: 99%

Sound-based Transportation Mode Recognition with Smartphones

Wang

Roggen

2019

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Smartphone-based identification of the mode of transportation of the user is important for context-aware services. We investigate the feasibility of recognizing the 8 most common modes of locomotion and transportation from the sound recorded by a smartphone carried by the user. We propose a convolutional neural network based recognition pipeline, which operates on the shorttime Fourier transform (STFT) spectrogram of the sound in the log domain. Experiment with the Sussex-Huawei locomotiontransportation (SHL) dataset on 366 hours of data shows promising results where the proposed pipeline can recognize the activities Still, Walk, Run, Bike, Car, Bus, Train and Subway with a global accuracy of 86.6%, which is 23% higher than classical machine learning pipelines. It is shown that sound is particularly useful for distinguishing between various vehicle activities (e.g. Car vs Bus, Train vs Subway). This discriminablity is complementary to the widely used motion sensors, which are poor at distinguish between rail and road transport.

show abstract

“…Sound-event classification has seen a noticeable increase in interest as a field of research [1][2][3][4][5][6]. Supported by advancements in information communication technology (ICT) and convergence technology in the Industry 4.0 era [7], various industries are conducting research using sound data.…”

Section: Introductionmentioning

confidence: 99%

Noise-Robust Sound-Event Classification System with Texture Analysis

et al. 2018

View full text Add to dashboard Cite

Sound-event classification has emerged as an important field of research in recent years. In particular, investigations using sound data are being conducted in various industrial fields. However, sound-event classification tasks have become more difficult and challenging with the increase in noise levels. In this study, we propose a noise-robust system for the classification of sound data. In this method, we first convert one-dimensional sound signals into two-dimensional gray-level images using normalization, and then extract the texture images by means of the dominant neighborhood structure (DNS) technique. Finally, we experimentally validate the noise-robust approach by using four classifiers (convolutional neural network (CNN), support vector machine (SVM), k-nearest neighbors(k-NN), and C4.5). The experimental results showed superior classification performance in noisy conditions compared with other methods. The F1 score exceeds 98.80% in railway data, and 96.57% in livestock data. Besides, the proposed method can be implemented in a cost-efficient manner (for instance, use of a low-cost microphone) while maintaining high level of accuracy in noisy environments. This approach can be used either as a standalone solution or as a supplement to the known methods to obtain a more accurate solution.

show abstract

Noise robust sound event classification with convolutional neural network

Cited by 85 publications

References 33 publications

Time–Frequency-Analysis-Based Blind Modulation Classification for Multiple-Antenna Systems

Time–Frequency-Analysis-Based Blind Modulation Classification for Multiple-Antenna Systems

Sound-based Transportation Mode Recognition with Smartphones

Noise-Robust Sound-Event Classification System with Texture Analysis

Contact Info

Product

Resources

About