Detection of impulsive sounds in stream of audio signals

Suliman, Azizah; Omarov, Batyrkhan; Dosbayev, Zhandos

doi:10.1109/icimu49871.2020.9243540

Cited by 7 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, we use a relatively simple neural network. In Table IV, SVM stands for Support Vector Machine [29], CNN stands for Convolutional Neural Network [30], and kNN for k-Nearest Neighbors [28]. V. CONCLUSIONS This paper presents a set of new features that have been developed for reliable acoustic detection of gunshots from hunting weapons in the wild.…”

Section: Achieved Resultsmentioning

confidence: 99%

Efficient Feature Set Developed for Acoustic Gunshot Detection in Open Space

Sigmund

Hrabina

2021

ELEKTRON ELEKTROTECH

View full text Add to dashboard Cite

This paper presents an efficient approach to automatic gunshot detection based on a combination of two feature sets: adapted standard sound features and hand-crafted novel features. The standard features are mel-frequency cepstral coefficients adapted for gunshot recognition in terms of uniform gamma-tone filters linearly spaced over the whole frequency range from 0 kHz to 16 kHz. The first 18 coefficients calculated from the 41 filters represent the best set of the optimized cepstral coefficients. The novel features were derived in the time domain from individual significant points of the raw waveform after amplitude normalization. Experiments were performed using single and ensemble neural networks to verify the effectiveness of the novel features for supplementing the standard features. The novelty of the work is the proposed feature combination, which allows to achieve very effective detection of gunshots from hunting weapons using 23 features and a simple neural network. In binary classification, the developed approach achieved an accuracy of 95.02 % in gunshot detection and 98.16 % in disregarding other sounds (i.e., non-gunshot).

show abstract

Section: Achieved Resultsmentioning

confidence: 99%

Efficient Feature Set Developed for Acoustic Gunshot Detection in Open Space

Sigmund

Hrabina

2021

ELEKTRON ELEKTROTECH

View full text Add to dashboard Cite

show abstract

“…In case of microphones arrays it is possible to use a two stage methodology comprising of a Blind System Identification and Deconvolution (BSID) stage followed by a SVM-based classification [19] for gunshot detection in a noisy urban environment. In [20] a method of classifying impulsive sounds based on a Weighted Majority Voting (WMV) strategy is described. In [21] Convolutional Neural Network (CNN) with temporal and spectral features is used for gunshot sound categories classification (pistol, rifle and shotgun of different calibres) reaching over 90 % accuracy.…”

Section: Related Workmentioning

confidence: 99%

Independent Channel Residual Convolutional Network for Gunshot Detection

Bajzík¹,

Přinosil²,

Jarina³

et al. 2022

IJACSA

View full text Add to dashboard Cite

The main purpose of this work is to propose a robust approach for dangerous sound events detection (e.g. gunshots) to improve recent surveillance systems. Despite the fact that the detection and classification of different sound events has a long history in signal processing, the analysis of environmental sounds is still challenging. The most recent works aim to prefer the time-frequency 2-D representation of sound as input to feed convolutional neural networks. This paper includes an analysis of known architectures as well as a newly proposed Independent Channel Residual Convolutional Network architecture based on standard residual blocks. Our approach consists of processing three different types of features in the individual channels. The UrbanSound8k and the Free Firearm Sound Library audio datasets are used for training and testing data generation, achieving a 98 % F1 score. The model was also evaluated in the wild using manually annotated movie audio track, achieving a 44 % F1 score, which is not too high but still better than other state-of-the-art techniques.

show abstract

Readout for simple and precise analog acoustic impact initialization

Šotner

Polák

Jeřábek

et al. 2021

Sci Rep

View full text Add to dashboard Cite

An economic concept of acoustic shock wave sensing readout system for simple computer processing is introduced in this work. Its application can be found in precise initialization of the stopwatch from the starter sound, handclap or gun in competitive sport races but also in many other places. The proposed device consists of several low-cost commercially available components and it is powered by a 9 V battery. The proposed device reliably reacts on incoming acoustic shock wave by generation of explicit impulse having controllable duration. It significantly overcomes basic implementations using only a microphone and amplifier (generating parasitic burst instead of defined and distinct impulse) or systems allowing a limited number of adjustable features (gain and/or threshold of the comparator—our concept offers the adjustment of gain, cut-off frequency, threshold level and time duration of active state). In comparison with standard methods, the proposed approach simplifies and makes sensing device less expensive and universal for any powder-based starting gun (without necessity to adapt starting gun). The proposed device, among others, has the following features: impulse duration can be controlled from hundreds of μs up to 2.3 s, the gain range of linear part of processing from 6 to 40 dB and open-collector output compatible with 5 V TTL or 3.3 V CMOS logic. The initialization has been tested in the range from tens of centimeters up to four meters. In order to highlight the important spectral components, the spectral character of the signal can be optimally reduced by a low-pass filter. The quiescent power consumption of the designed simple analog circuit reaches 90 mW. Several use cases, response of the designed system on gunshot signature, talking, hand-clapping and hit on the sensing microphone, are studied and compared to each other. Simulation and experimental results confirm functionality of the realized system.

show abstract

Detection of impulsive sounds in stream of audio signals

Cited by 7 publications

References 20 publications

Efficient Feature Set Developed for Acoustic Gunshot Detection in Open Space

Efficient Feature Set Developed for Acoustic Gunshot Detection in Open Space

Independent Channel Residual Convolutional Network for Gunshot Detection

Readout for simple and precise analog acoustic impact initialization

Contact Info

Product

Resources

About