This paper focuses on an experimental study that used passive sonar sensors as the primary information
source for the submerged target in order to identify, classify, and recognize naval targets. Surface vessels
and submarine generate a specific sound either by propulsion systems, auxiliary equipment or blades of
their propellers, producing information known as the "acoustic signature" that is unique to each type of
target. Consequently, the analysis and classification of targets depend on the processing of the frequencies
produced by these vibrations (sound). utilizing the TPWS (Two-Pass-Split Windows) filter, this work aims
to develop a novel technique for target identification and classification utilizing passive sonars. This
technique involves processing the target's signal in the time-frequency domain. subsequently, in order to
improve the frequency lines of the target noise and decrease the background noise, a TPSW algorithm is
implemented in the frequency domain. By integrating narrowband and broadband analysis as inputs of an
artificial intelligence model that can classify a target into one of the categories given in the training phase,
the target has finally been classified. Our findings demonstrated that the suggested approach is dependent
upon the size of the target noise data collection and the noise-to-effective-signal ratio.
SHIPS MATCHING BASED ON AN ADAPTIVE ACOUSTIC SPECTRUM SIGNATURE DETECTION ALG...sipij
In this paper, an acoustic spectrum signature tracks matching algorithm based on the Manhattan distance and the Euclidean distance of signature vectors, and a multi-frame fusion algorithm are proposed for reliable real time detection and matching of boat generated acoustic signal spectrum signatures. The
experiments results have shown that the proposed tracks matching algorithm has the ability to discriminate the tracks from different ships and the ability of matching of the tracks from the same ship; and the spectrum signature detection algorithm has captured the critical features of ship generated acoustic
signals. In the process of signal spectrum signature detection, the observation of time and frequency space is structured by dividing input digitalized acoustic signal into multiple frames and each frame is transformed into the frequency domain by FFT. Then, a normalization of signal spectrum vector is carried
out to make the detection process more robust. After that, an adaptive median Constant False Alarm Rate (AMCFAR) algorithm is used for the detection and extraction of boat generated spectrum signature, in which an extreme low constant false alarm rate is kept with relative high detection rate. Finally, the frame detections are accumulated to build up the track spectrum signatures.
Spectrum Sensing Detection with Sequential Forward Search in Comparison to Kn...IJMTST Journal
FCC is currently working on the concept of white space users “borrowing” spectrum from free license
holders temporarily to improve the spectrum utilization.
This project provides a relation between a Pf and the SNR value of any spectrum detector to have a
certain performance. Previous spectrum sensing detection techniques are only suitable for Low SNR and
are based on signal information values. But these methods are purely narrow band spectrum applications
In order to overcome the above said drawbacks we propose a novel method of spectrum sensing method
and is suitable for low and high SNR values, the sensed spectrum applicable for wide band applications.
Our proposed method does not require signal information at the receiver and channel information, because
this flexibility sensing rate is very high compared to previous techniques.
Development of an FHMA-based Underwater Acoustic Communications System for Mu...Waqas Tariq
This paper describes the design of an underwater acoustic communications system for multiple underwater vehicles, based on frequency-hopping multiple-access (FHMA) and tamed spread-spectrum communications. The system makes used of the tamed spread-spectrum method, frequency hopping, 4FSK, and a rake receiver. In order to make the system more practical, the underwater channel and the effect of the number of users on the bit error ratio (BER) are also taken into account. Since the necessary proving experiments are not easily conducted in the ocean, a platform is developed that uses the sound card of a computer, combined with a sound box and microphone, to transduce energy for acoustic communications. Simulated and experimental results indicate that this system could provide reliable underwater communications between multiple underwater vehicles.
MULTI-STAGES CO-OPERATIVE/NONCOOPERATIVE SCHEMES OF SPECTRUM SENSING FOR COGN...ijwmn
Searching for spectrum holes in practical wireless channels where primary users experience multipath
fading and shadowing, with noise uncertainty, limits the detection performance significantly. Moreover, the
detection challenge will be tougher when different band types have to be sensed, with different signal and
spectral characteristics, and probably overlapping spectra. Besides, primary user waveforms can be known
(completely or partially) or unknown to allow or forbid cognitive radios to use specific kinds of detection
schemes! Hidden primary user’s problem, and doubly selective channel oblige the use of cooperative
sensing to exploit the spatial diversity in the observations of spatially located cognitive radio users.
Incorporated all the aforementioned practical challenges as a whole, this paper developed a new multistage detection scheme that intelligently decides the detection algorithm based on power, noise, bandwidth
and knowledge of the signal of interest. The proposed scheme switches between individual and cooperative
sensing and among featured based sensing techniques (cyclo-stationary detection and matched filter) and
sub-band energy detection according to the characteristics of signal and band of interest.Compared to the
existing schemes, performance evaluations show reliable results in terms of probabilities of detection and
mean sensing times under the aforementioned conditions.
MULTI-STAGES CO-OPERATIVE/NONCOOPERATIVE SCHEMES OF SPECTRUM SENSING FOR COGN...ijwmn
Searching for spectrum holes in practical wireless channels where primary users experience multipath
fading and shadowing, with noise uncertainty, limits the detection performance significantly. Moreover, the
detection challenge will be tougher when different band types have to be sensed, with different signal and
spectral characteristics, and probably overlapping spectra. Besides, primary user waveforms can be known
(completely or partially) or unknown to allow or forbid cognitive radios to use specific kinds of detection
schemes! Hidden primary user’s problem, and doubly selective channel oblige the use of cooperative
sensing to exploit the spatial diversity in the observations of spatially located cognitive radio users.
Incorporated all the aforementioned practical challenges as a whole, this paper developed a new multistage detection scheme that intelligently decides the detection algorithm based on power, noise, bandwidth
and knowledge of the signal of interest. The proposed scheme switches between individual and cooperative
sensing and among featured based sensing techniques (cyclo-stationary detection and matched filter) and
sub-band energy detection according to the characteristics of signal and band of interest.Compared to the
existing schemes, performance evaluations show reliable results in terms of probabilities of detection and
mean sensing times under the aforementioned conditions.
SHIPS MATCHING BASED ON AN ADAPTIVE ACOUSTIC SPECTRUM SIGNATURE DETECTION ALG...sipij
In this paper, an acoustic spectrum signature tracks matching algorithm based on the Manhattan distance and the Euclidean distance of signature vectors, and a multi-frame fusion algorithm are proposed for reliable real time detection and matching of boat generated acoustic signal spectrum signatures. The
experiments results have shown that the proposed tracks matching algorithm has the ability to discriminate the tracks from different ships and the ability of matching of the tracks from the same ship; and the spectrum signature detection algorithm has captured the critical features of ship generated acoustic
signals. In the process of signal spectrum signature detection, the observation of time and frequency space is structured by dividing input digitalized acoustic signal into multiple frames and each frame is transformed into the frequency domain by FFT. Then, a normalization of signal spectrum vector is carried
out to make the detection process more robust. After that, an adaptive median Constant False Alarm Rate (AMCFAR) algorithm is used for the detection and extraction of boat generated spectrum signature, in which an extreme low constant false alarm rate is kept with relative high detection rate. Finally, the frame detections are accumulated to build up the track spectrum signatures.
Spectrum Sensing Detection with Sequential Forward Search in Comparison to Kn...IJMTST Journal
FCC is currently working on the concept of white space users “borrowing” spectrum from free license
holders temporarily to improve the spectrum utilization.
This project provides a relation between a Pf and the SNR value of any spectrum detector to have a
certain performance. Previous spectrum sensing detection techniques are only suitable for Low SNR and
are based on signal information values. But these methods are purely narrow band spectrum applications
In order to overcome the above said drawbacks we propose a novel method of spectrum sensing method
and is suitable for low and high SNR values, the sensed spectrum applicable for wide band applications.
Our proposed method does not require signal information at the receiver and channel information, because
this flexibility sensing rate is very high compared to previous techniques.
Development of an FHMA-based Underwater Acoustic Communications System for Mu...Waqas Tariq
This paper describes the design of an underwater acoustic communications system for multiple underwater vehicles, based on frequency-hopping multiple-access (FHMA) and tamed spread-spectrum communications. The system makes used of the tamed spread-spectrum method, frequency hopping, 4FSK, and a rake receiver. In order to make the system more practical, the underwater channel and the effect of the number of users on the bit error ratio (BER) are also taken into account. Since the necessary proving experiments are not easily conducted in the ocean, a platform is developed that uses the sound card of a computer, combined with a sound box and microphone, to transduce energy for acoustic communications. Simulated and experimental results indicate that this system could provide reliable underwater communications between multiple underwater vehicles.
MULTI-STAGES CO-OPERATIVE/NONCOOPERATIVE SCHEMES OF SPECTRUM SENSING FOR COGN...ijwmn
Searching for spectrum holes in practical wireless channels where primary users experience multipath
fading and shadowing, with noise uncertainty, limits the detection performance significantly. Moreover, the
detection challenge will be tougher when different band types have to be sensed, with different signal and
spectral characteristics, and probably overlapping spectra. Besides, primary user waveforms can be known
(completely or partially) or unknown to allow or forbid cognitive radios to use specific kinds of detection
schemes! Hidden primary user’s problem, and doubly selective channel oblige the use of cooperative
sensing to exploit the spatial diversity in the observations of spatially located cognitive radio users.
Incorporated all the aforementioned practical challenges as a whole, this paper developed a new multistage detection scheme that intelligently decides the detection algorithm based on power, noise, bandwidth
and knowledge of the signal of interest. The proposed scheme switches between individual and cooperative
sensing and among featured based sensing techniques (cyclo-stationary detection and matched filter) and
sub-band energy detection according to the characteristics of signal and band of interest.Compared to the
existing schemes, performance evaluations show reliable results in terms of probabilities of detection and
mean sensing times under the aforementioned conditions.
MULTI-STAGES CO-OPERATIVE/NONCOOPERATIVE SCHEMES OF SPECTRUM SENSING FOR COGN...ijwmn
Searching for spectrum holes in practical wireless channels where primary users experience multipath
fading and shadowing, with noise uncertainty, limits the detection performance significantly. Moreover, the
detection challenge will be tougher when different band types have to be sensed, with different signal and
spectral characteristics, and probably overlapping spectra. Besides, primary user waveforms can be known
(completely or partially) or unknown to allow or forbid cognitive radios to use specific kinds of detection
schemes! Hidden primary user’s problem, and doubly selective channel oblige the use of cooperative
sensing to exploit the spatial diversity in the observations of spatially located cognitive radio users.
Incorporated all the aforementioned practical challenges as a whole, this paper developed a new multistage detection scheme that intelligently decides the detection algorithm based on power, noise, bandwidth
and knowledge of the signal of interest. The proposed scheme switches between individual and cooperative
sensing and among featured based sensing techniques (cyclo-stationary detection and matched filter) and
sub-band energy detection according to the characteristics of signal and band of interest.Compared to the
existing schemes, performance evaluations show reliable results in terms of probabilities of detection and
mean sensing times under the aforementioned conditions.
Marine Vehicle Spectrum Signature Detection Based On An Adaptive CFAR and Mul...CSCJournals
Detecting marine vehicle spectrum signature from hydrophone at low false alarm rate and high detection rate in an environment of various interference is a very difficult problem. To overcome this problem, an observation space is created by sampling and dividing input analog acoustic signal into digital signal in multiple frames and each frame is transformed into the frequency domain; then an Adaptive Constant False Alarm Rate (ACFAR) and Post Detection Fusion algorithms have been proposed for an effective automatic detection of marine vehicle generated acoustic signal spectrum signature. The proposed algorithms have been tested on several real acoustic signals. The statistical analysis and experimental results showed that the proposed algorithm has kept a very low false alarm rate and extremely high detection rate.
Clutter reduction technique based on clutter model for automatic target class...TELKOMNIKA JOURNAL
Classification becomes one of the important elements in the forward scatter radar (FSR) micro-sensors network. This classification performance is dependent on the target’s profile behaviour and the network’s surrounding; and one of the factors that cause the reduction of classification probability is the presence of ground clutter. As the volume of clutter increases, their masking effect becomes greater and may result in more significant errors in target classification. Hence, to reduce misclassification in the FSR sensor network, a new clutter reduction technique based on the ground clutter model is proposed. Simulated ground clutter is modeled based on the estimated signal to clutter ratio (SCR) of the received signal. The clutter effect is diminished by eliminating simulated like-clutter from the receiving signals. The result shows improvement in the classification accuracy, especially for the minimum value of the SCR and this new technique uses only one database which will shorten the processing time and reduce the overall database’s size.
Audio Features Based Steganography Detection in WAV Fileijtsrd
Audio signals containing secret information or not is a security issue addressed in the context of steganalysis. ThRainfalle conceptual ide lies in the difference of the distribution of various statistical distance measures between the cover audio signals and stego audio signals. The aim of the propose system is to analyze the audio signal which have the presence of information hiding behavior or not. Mel frequency ceptral coefficient, zero crossing rate, spectral flux and short time energy features of audio signal are extracted, and combine these features with the features extracted from the modified version that is generated by randomly modifying with significant bits. Moreover, the extracted features are detected or classified with a support vector machine in this propose system. Experimental result show that the propose method performs well in steganalysis of the audio stegnograms that are produced by using S tools4. Khin Myo Kyi "Audio Features Based Steganography Detection in WAV File" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-5 , August 2019, URL: https://www.ijtsrd.com/papers/ijtsrd26807.pdf Paper URL: https://www.ijtsrd.com/computer-science/other/26807/audio-features-based-steganography-detection-in-wav-file/khin-myo-kyi
Error Rate Performance of Interleaved Coded OFDM For Undersea Acoustic LinksCSCJournals
Studies on undersea acoustic communication links, set up through highly complex and inhomogeneous underwater channel using various orders of QAM and PSK based OFDM techniques, have been reported in open literature. However, their bit error rate performances still need to be improved. Coding, when combined with OFDM, helps to detect and correct errors without having the overhead of too many retransmissions, as the bandwidth is a scarce resource in undersea scenario. The technique of interleaving, which is frequently employed in digital communication and storage systems to enhance the performance of the coding schemes, can be used to improve the error rate performance of the coded OFDM. The error rate performances of interleaved convolutional and BCH coded OFDMs for undersea acoustic links for binary phase shift keying and its differential variant have been studied in this paper. It is found that at high SNR, the process of interleaving and coding offers significant improvement in the error rate performance. It is also worth mentioning the fact that interleaving improves the performance of both convolutional and BCH coded OFDM systems.
Energy Efficient Animal Sound Recognition Scheme in Wireless Acoustic Sensors...ijwmn
Wireless sensor network (WSN) has proliferated rapidly as a cost-effective solution for data aggregation and measurements under challenging environments. Sensors in WSNs are cheap, powerful, and consume limited energy. The energy consumption is considered to be the dominant concern because it has a direct and significant influence on the application’s lifetime. Recently, the availability of small and inexpensive components such as microphones has promoted the development of wireless acoustic sensor networks (WASNs). Examples of WASN applications are hearing aids, acoustic monitoring, and ambient intelligence. Monitoring animals, especially those that are becoming endangered, can assist with biology researchers’ preservation efforts. In this work, we first focus on exploring the existing methods used to monitor the animal by recognizing their sounds. Then we propose a new energy-efficient approach for identifying animal sounds based on the frequency features extracted from acoustic sensed data. This approach represents a suitable solution that can be implemented and used in various applications. However, the proposed system considers the balance between application efficiency and the sensor’s energy capabilities. The energy savings will be achieved through processing the recognition tasks in each sensor, and the recognition results will be sent to the base station.
ENERGY EFFICIENT ANIMAL SOUND RECOGNITION SCHEME IN WIRELESS ACOUSTIC SENSORS...ijwmn
Wireless sensor network (WSN) has proliferated rapidly as a cost-effective solution for data aggregation and measurements under challenging environments. Sensors in WSNs are cheap, powerful, and consume limited energy. The energy consumption is considered to be the dominant concern because it has a direct and significant influence on the application’s lifetime. Recently, the availability of small and inexpensive components such as microphones has promoted the development of wireless acoustic sensor networks (WASNs). Examples of WASN applications are hearing aids, acoustic monitoring, and ambient intelligence. Monitoring animals, especially those that are becoming endangered, can assist with biology researchers’ preservation efforts. In this work, we first focus on exploring the existing methods used to monitor the animal by recognizing their sounds. Then we propose a new energy-efficient approach for identifying animal sounds based on the frequency features extracted from acoustic sensed data. This approach represents a suitable solution that can be implemented and used in various applications. However, the proposed system considers the balance between application efficiency and the sensor’s energy capabilities. The energy savings will be achieved through processing the recognition tasks in each sensor, and the recognition results will be sent to the base station.
NOISE REMOVAL TECHNIQUES FOR MICROWAVE REMOTE SENSING RADAR DATA AND ITS EVAL...cscpconf
Microwave Remote Sensing data acquired by a RADAR sensor such as SAR(Synthetic Aperture Radar) is affected by a peculiar kind of noise called speckle. This noise not only renders the
data ineffective for classification, texture analysis, segmentation etc. which are used for image analysis purposes, but also degrades the overall contrast and radiometric quality of the image. Here we discuss the various noise removal techniques which have been widely used by scientists all over the world. Different filtering methods have their pros and cons, and no single method can give the most satisfactory result. In order to circumvent those issues, better and better methods are being attempted. One of the recent methods is that based on Wavelet technique. This paper discusses the denoising techniques based on Wavelets and the results from some of those methods. The relative merits and demerits of the filters and their evaluation is also done.
Noise removal techniques for microwave remote sensing radar data and its eval...csandit
Microwave Remote Sensing data acquired by a RADAR sensor such as SAR(Synthetic Aperture
Radar) is affected by a peculiar kind of noise called speckle. This noise not only renders the
data ineffective for classification, texture analysis, segmentation etc. which are used for image
analysis purposes, but also degrades the overall contrast and radiometric quality of the image.
Here we discuss the various noise removal techniques which have been widely used by scientists
all over the world. Different filtering methods have their pros and cons, and no single method
can give the most satisfactory result. In order to circumvent those issues, better and better
methods are being attempted. One of the recent methods is that based on Wavelet technique.
This paper discusses the denoising techniques based on Wavelets and the results from some of
those methods. The relative merits and demerits of the filters and their evaluation is also done.
A Gaussian Clustering Based Voice Activity Detector for Noisy Environments Us...CSCJournals
In this paper, a voice activity detector is proposed on the basis of Gaussian modeling of noise in the spectro-temporal space. Spectro-temporal space is obtained from auditory cortical processing. The auditory model that offers a multi-dimensional picture of the sound includes two stages: the initial stage is a model of inner ear and the second stage is the auditory central cortical modeling in the brain. In this paper, the speech noise in this picture has been modeled by a 3-D mono Gaussian cluster. At the start of suggested VAD process, the noise is modeled by a Gaussian shaped cluster. The average noise behavior is obtained in different spectrotemporal space in various points for each frame. In the stage of separation of speech from noise, the criterion is the difference between the average noise behavior and the speech signal amplitude in spectrotemporal domain. This was measured for each frame and was used as the criterion of classification. Using Noisex92, this method is tested in different noise models such as White, exhibition, Street, Office and Train noises. The results are compared to both auditory model and multifeature method. It is observed that the performance of this method in low signal-to-noise ratios (SNRs) conditions is better than other current methods.
Sonar, (from “sound navigation ranging”), technique for detecting and determining the distance and direction of underwater objects by acoustic means. Sound waves emitted by or reflected from the object are detected by sonar apparatus and analyzed for the information they contain.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
Immunizing Image Classifiers Against Localized Adversary Attacksgerogepatton
This paper addresses the vulnerability of deep learning models, particularly convolutional neural networks
(CNN)s, to adversarial attacks and presents a proactive training technique designed to counter them. We
introduce a novel volumization algorithm, which transforms 2D images into 3D volumetric representations.
When combined with 3D convolution and deep curriculum learning optimization (CLO), itsignificantly improves
the immunity of models against localized universal attacks by up to 40%. We evaluate our proposed approach
using contemporary CNN architectures and the modified Canadian Institute for Advanced Research (CIFAR-10
and CIFAR-100) and ImageNet Large Scale Visual Recognition Challenge (ILSVRC12) datasets, showcasing
accuracy improvements over previous techniques. The results indicate that the combination of the volumetric
input and curriculum learning holds significant promise for mitigating adversarial attacks without necessitating
adversary training.
More Related Content
Similar to Passive Sonar Detection and Classification Based on Demon-Lofar Analysis and Neural Network Algorithms
Marine Vehicle Spectrum Signature Detection Based On An Adaptive CFAR and Mul...CSCJournals
Detecting marine vehicle spectrum signature from hydrophone at low false alarm rate and high detection rate in an environment of various interference is a very difficult problem. To overcome this problem, an observation space is created by sampling and dividing input analog acoustic signal into digital signal in multiple frames and each frame is transformed into the frequency domain; then an Adaptive Constant False Alarm Rate (ACFAR) and Post Detection Fusion algorithms have been proposed for an effective automatic detection of marine vehicle generated acoustic signal spectrum signature. The proposed algorithms have been tested on several real acoustic signals. The statistical analysis and experimental results showed that the proposed algorithm has kept a very low false alarm rate and extremely high detection rate.
Clutter reduction technique based on clutter model for automatic target class...TELKOMNIKA JOURNAL
Classification becomes one of the important elements in the forward scatter radar (FSR) micro-sensors network. This classification performance is dependent on the target’s profile behaviour and the network’s surrounding; and one of the factors that cause the reduction of classification probability is the presence of ground clutter. As the volume of clutter increases, their masking effect becomes greater and may result in more significant errors in target classification. Hence, to reduce misclassification in the FSR sensor network, a new clutter reduction technique based on the ground clutter model is proposed. Simulated ground clutter is modeled based on the estimated signal to clutter ratio (SCR) of the received signal. The clutter effect is diminished by eliminating simulated like-clutter from the receiving signals. The result shows improvement in the classification accuracy, especially for the minimum value of the SCR and this new technique uses only one database which will shorten the processing time and reduce the overall database’s size.
Audio Features Based Steganography Detection in WAV Fileijtsrd
Audio signals containing secret information or not is a security issue addressed in the context of steganalysis. ThRainfalle conceptual ide lies in the difference of the distribution of various statistical distance measures between the cover audio signals and stego audio signals. The aim of the propose system is to analyze the audio signal which have the presence of information hiding behavior or not. Mel frequency ceptral coefficient, zero crossing rate, spectral flux and short time energy features of audio signal are extracted, and combine these features with the features extracted from the modified version that is generated by randomly modifying with significant bits. Moreover, the extracted features are detected or classified with a support vector machine in this propose system. Experimental result show that the propose method performs well in steganalysis of the audio stegnograms that are produced by using S tools4. Khin Myo Kyi "Audio Features Based Steganography Detection in WAV File" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-5 , August 2019, URL: https://www.ijtsrd.com/papers/ijtsrd26807.pdf Paper URL: https://www.ijtsrd.com/computer-science/other/26807/audio-features-based-steganography-detection-in-wav-file/khin-myo-kyi
Error Rate Performance of Interleaved Coded OFDM For Undersea Acoustic LinksCSCJournals
Studies on undersea acoustic communication links, set up through highly complex and inhomogeneous underwater channel using various orders of QAM and PSK based OFDM techniques, have been reported in open literature. However, their bit error rate performances still need to be improved. Coding, when combined with OFDM, helps to detect and correct errors without having the overhead of too many retransmissions, as the bandwidth is a scarce resource in undersea scenario. The technique of interleaving, which is frequently employed in digital communication and storage systems to enhance the performance of the coding schemes, can be used to improve the error rate performance of the coded OFDM. The error rate performances of interleaved convolutional and BCH coded OFDMs for undersea acoustic links for binary phase shift keying and its differential variant have been studied in this paper. It is found that at high SNR, the process of interleaving and coding offers significant improvement in the error rate performance. It is also worth mentioning the fact that interleaving improves the performance of both convolutional and BCH coded OFDM systems.
Energy Efficient Animal Sound Recognition Scheme in Wireless Acoustic Sensors...ijwmn
Wireless sensor network (WSN) has proliferated rapidly as a cost-effective solution for data aggregation and measurements under challenging environments. Sensors in WSNs are cheap, powerful, and consume limited energy. The energy consumption is considered to be the dominant concern because it has a direct and significant influence on the application’s lifetime. Recently, the availability of small and inexpensive components such as microphones has promoted the development of wireless acoustic sensor networks (WASNs). Examples of WASN applications are hearing aids, acoustic monitoring, and ambient intelligence. Monitoring animals, especially those that are becoming endangered, can assist with biology researchers’ preservation efforts. In this work, we first focus on exploring the existing methods used to monitor the animal by recognizing their sounds. Then we propose a new energy-efficient approach for identifying animal sounds based on the frequency features extracted from acoustic sensed data. This approach represents a suitable solution that can be implemented and used in various applications. However, the proposed system considers the balance between application efficiency and the sensor’s energy capabilities. The energy savings will be achieved through processing the recognition tasks in each sensor, and the recognition results will be sent to the base station.
ENERGY EFFICIENT ANIMAL SOUND RECOGNITION SCHEME IN WIRELESS ACOUSTIC SENSORS...ijwmn
Wireless sensor network (WSN) has proliferated rapidly as a cost-effective solution for data aggregation and measurements under challenging environments. Sensors in WSNs are cheap, powerful, and consume limited energy. The energy consumption is considered to be the dominant concern because it has a direct and significant influence on the application’s lifetime. Recently, the availability of small and inexpensive components such as microphones has promoted the development of wireless acoustic sensor networks (WASNs). Examples of WASN applications are hearing aids, acoustic monitoring, and ambient intelligence. Monitoring animals, especially those that are becoming endangered, can assist with biology researchers’ preservation efforts. In this work, we first focus on exploring the existing methods used to monitor the animal by recognizing their sounds. Then we propose a new energy-efficient approach for identifying animal sounds based on the frequency features extracted from acoustic sensed data. This approach represents a suitable solution that can be implemented and used in various applications. However, the proposed system considers the balance between application efficiency and the sensor’s energy capabilities. The energy savings will be achieved through processing the recognition tasks in each sensor, and the recognition results will be sent to the base station.
NOISE REMOVAL TECHNIQUES FOR MICROWAVE REMOTE SENSING RADAR DATA AND ITS EVAL...cscpconf
Microwave Remote Sensing data acquired by a RADAR sensor such as SAR(Synthetic Aperture Radar) is affected by a peculiar kind of noise called speckle. This noise not only renders the
data ineffective for classification, texture analysis, segmentation etc. which are used for image analysis purposes, but also degrades the overall contrast and radiometric quality of the image. Here we discuss the various noise removal techniques which have been widely used by scientists all over the world. Different filtering methods have their pros and cons, and no single method can give the most satisfactory result. In order to circumvent those issues, better and better methods are being attempted. One of the recent methods is that based on Wavelet technique. This paper discusses the denoising techniques based on Wavelets and the results from some of those methods. The relative merits and demerits of the filters and their evaluation is also done.
Noise removal techniques for microwave remote sensing radar data and its eval...csandit
Microwave Remote Sensing data acquired by a RADAR sensor such as SAR(Synthetic Aperture
Radar) is affected by a peculiar kind of noise called speckle. This noise not only renders the
data ineffective for classification, texture analysis, segmentation etc. which are used for image
analysis purposes, but also degrades the overall contrast and radiometric quality of the image.
Here we discuss the various noise removal techniques which have been widely used by scientists
all over the world. Different filtering methods have their pros and cons, and no single method
can give the most satisfactory result. In order to circumvent those issues, better and better
methods are being attempted. One of the recent methods is that based on Wavelet technique.
This paper discusses the denoising techniques based on Wavelets and the results from some of
those methods. The relative merits and demerits of the filters and their evaluation is also done.
A Gaussian Clustering Based Voice Activity Detector for Noisy Environments Us...CSCJournals
In this paper, a voice activity detector is proposed on the basis of Gaussian modeling of noise in the spectro-temporal space. Spectro-temporal space is obtained from auditory cortical processing. The auditory model that offers a multi-dimensional picture of the sound includes two stages: the initial stage is a model of inner ear and the second stage is the auditory central cortical modeling in the brain. In this paper, the speech noise in this picture has been modeled by a 3-D mono Gaussian cluster. At the start of suggested VAD process, the noise is modeled by a Gaussian shaped cluster. The average noise behavior is obtained in different spectrotemporal space in various points for each frame. In the stage of separation of speech from noise, the criterion is the difference between the average noise behavior and the speech signal amplitude in spectrotemporal domain. This was measured for each frame and was used as the criterion of classification. Using Noisex92, this method is tested in different noise models such as White, exhibition, Street, Office and Train noises. The results are compared to both auditory model and multifeature method. It is observed that the performance of this method in low signal-to-noise ratios (SNRs) conditions is better than other current methods.
Sonar, (from “sound navigation ranging”), technique for detecting and determining the distance and direction of underwater objects by acoustic means. Sound waves emitted by or reflected from the object are detected by sonar apparatus and analyzed for the information they contain.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
Immunizing Image Classifiers Against Localized Adversary Attacksgerogepatton
This paper addresses the vulnerability of deep learning models, particularly convolutional neural networks
(CNN)s, to adversarial attacks and presents a proactive training technique designed to counter them. We
introduce a novel volumization algorithm, which transforms 2D images into 3D volumetric representations.
When combined with 3D convolution and deep curriculum learning optimization (CLO), itsignificantly improves
the immunity of models against localized universal attacks by up to 40%. We evaluate our proposed approach
using contemporary CNN architectures and the modified Canadian Institute for Advanced Research (CIFAR-10
and CIFAR-100) and ImageNet Large Scale Visual Recognition Challenge (ILSVRC12) datasets, showcasing
accuracy improvements over previous techniques. The results indicate that the combination of the volumetric
input and curriculum learning holds significant promise for mitigating adversarial attacks without necessitating
adversary training.
May 2024 - Top 10 Read Articles in Artificial Intelligence and Applications (...gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
3rd International Conference on Artificial Intelligence Advances (AIAD 2024)gerogepatton
3rd International Conference on Artificial Intelligence Advances (AIAD 2024) will act as a major forum for the presentation of innovative ideas, approaches, developments, and research projects in the area advanced Artificial Intelligence. It will also serve to facilitate the exchange of information between researchers and industry professionals to discuss the latest issues and advancement in the research area. Core areas of AI and advanced multi-disciplinary and its applications will be covered during the conferences.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
Information Extraction from Product Labels: A Machine Vision Approachgerogepatton
This research tackles the challenge of manual data extraction from product labels by employing a blend of
computer vision and Natural Language Processing (NLP). We introduce an enhanced model that combines
Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) in a Convolutional
Recurrent Neural Network (CRNN) for reliable text recognition. Our model is further refined by
incorporating the Tesseract OCR engine, enhancing its applicability in Optical Character Recognition
(OCR) tasks. The methodology is augmented by NLP techniques and extended through the Open Food
Facts API (Application Programming Interface) for database population and text-only label prediction.
The CRNN model is trained on encoded labels and evaluated for accuracy on a dedicated test set.
Importantly, our approach enables visually impaired individuals to access essential information on
product labels, such as directions and ingredients. Overall, the study highlights the efficacy of deep
learning and OCR in automating label extraction and recognition.
10th International Conference on Artificial Intelligence and Applications (AI...gerogepatton
10th International Conference on Artificial Intelligence and Applications (AI 2024) will provide an excellent international forum for sharing knowledge and results in theory, methodology and applications of Artificial Intelligence and its applications. The Conference looks for significant contributions to all major fields of the Artificial Intelligence, Soft Computing in theoretical and practical aspects. The aim of the Conference is to provide a platform to the researchers and practitioners from both academia as well as industry to meet and share cutting-edge development in the field.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
Research on Fuzzy C- Clustering Recursive Genetic Algorithm based on Cloud Co...gerogepatton
Aiming at the problems of poor local search ability and precocious convergence of fuzzy C-cluster
recursive genetic algorithm (FOLD++), a new fuzzy C-cluster recursive genetic algorithm based on
Bayesian function adaptation search (TS) was proposed by incorporating the idea of Bayesian function
adaptation search into fuzzy C-cluster recursive genetic algorithm. The new algorithm combines the
advantages of FOLD++ and TS. In the early stage of optimization, fuzzy C-cluster recursive genetic
algorithm is used to get a good initial value, and the individual extreme value pbest is put into Bayesian
function adaptation table. In the late stage of optimization, when the searching ability of fuzzy C-cluster
recursive genetic is weakened, the short term memory function of Bayesian function adaptation table in
Bayesian function adaptation search algorithm is utilized. Make it jump out of the local optimal solution,
and allow bad solutions to be accepted during the search. The improved algorithm is applied to function
optimization, and the simulation results show that the calculation accuracy and stability of the algorithm
are improved, and the effectiveness of the improved algorithm is verified
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
10th International Conference on Artificial Intelligence and Soft Computing (...gerogepatton
10th International Conference on Artificial Intelligence and Soft Computing (AIS 2024) will
provide an excellent international forum for sharing knowledge and results in theory, methodology, and
applications of Artificial Intelligence, Soft Computing. The Conference looks for significant
contributions to all major fields of the Artificial Intelligence, Soft Computing in theoretical and practical
aspects. The aim of the Conference is to provide a platform to the researchers and practitioners from
both academia as well as industry to meet and share cutting-edge development in the field.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
Employee attrition refers to the decrease in staff numbers within an organization due to various reasons.
As it has a negative impact on long-term growth objectives and workplace productivity, firms have
recognized it as a significant concern. To address this issue, organizations are increasingly turning to
machine-learning approaches to forecast employee attrition rates. This topic has gained significant
attention from researchers, especially in recent times. Several studies have applied various machinelearning methods to predict employee attrition, producing different resultsdepending on the employed
methods, factors, and datasets. However, there has been no comprehensive comparative review of multiple
studies applying machine-learning models to predict employee attrition to date. Therefore, this study aims
to fill this gap by providing an overview of research conducted on applying machine learning to predict
employee attrition from 2019 to February 2024. A literature review of relevant studies was conducted,
summarized, and classified. Most studies agree on conducting comparative experiments with multiple
predictive models to determine the most effective one.From this literature survey, the RF algorithm and
XGB ensemble method are repeatedly the best-performing, outperforming many other algorithms.
Additionally, the application of deep learning to employee attrition prediction issues also shows promise.
While there are discrepancies in the datasets used in previous studies, it is notable that the dataset
provided by IBM is the most widely utilized. This study serves as a concise review for new researchers,
facilitating their understanding of the primary techniques employed in predicting employee attrition and
highlighting recent research trends in this field. Furthermore, it provides organizations with insight into
the prominent factors affecting employee attrition, as identified by studies, enabling them to implement
solutions aimed at reducing attrition rates.
10th International Conference on Artificial Intelligence and Applications (AI...gerogepatton
10th International Conference on Artificial Intelligence and Applications (AIFU 2024) is a forum for presenting new advances and research results in the fields of Artificial Intelligence. The conference will bring together leading researchers, engineers and scientists in the domain of interest from around the world. The scope of the conference covers all theoretical and practical aspects of the Artificial Intelligence.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
THE TRANSFORMATION RISK-BENEFIT MODEL OF ARTIFICIAL INTELLIGENCE:BALANCING RI...gerogepatton
This paper summarizes the most cogent advantages and risks associated with Artificial Intelligence from an
in-depth review of the literature. Then the authors synthesize the salient risk-related models currently being
used in AI, technology and business-related scenarios. Next, in view of an updated context of AI along with
theories and models reviewed and expanded constructs, the writers propose a new framework called “The
Transformation Risk-Benefit Model of Artificial Intelligence” to address the increasing fears and levels of
AIrisk. Using the model characteristics, the article emphasizes practical and innovative solutions where
benefitsoutweigh risks and three use cases in healthcare, climate change/environment and cyber security to
illustrate unique interplay of principles, dimensions and processes of this powerful AI transformational
model.
13th International Conference on Software Engineering and Applications (SEA 2...gerogepatton
13th International Conference on Software Engineering and Applications (SEA 2024) will provide an excellent international forum for sharing knowledge and results in theory, methodology and applications of Software Engineering and Applications. The goal of this conference is to bring together researchers and practitioners from academia and industry to focus on understanding Modern software engineering concepts and establishing new collaborations in these areas.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
AN IMPROVED MT5 MODEL FOR CHINESE TEXT SUMMARY GENERATIONgerogepatton
Complicated policy texts require a lot of effort to read, so there is a need for intelligent interpretation of
Chinese policies. To better solve the Chinese Text Summarization task, this paper utilized the mT5 model
as the core framework and initial weights. Additionally, In addition, this paper reduced the model size
through parameter clipping, used the Gap Sentence Generation (GSG) method as unsupervised method,
and improved the Chinese tokenizer. After training on a meticulously processed 30GB Chinese training
corpus, the paper developed the enhanced mT5-GSG model. Then, when fine-tuning the Chinese Policy
text, this paper chose the idea of “Dropout Twice”, and innovatively combined the probability distribution
of the two Dropouts through the Wasserstein distance. Experimental results indicate that the proposed
model achieved Rouge-1, Rouge-2, and Rouge-L scores of 56.13%, 45.76%, and 56.41% respectively on
the Chinese policy text summarization dataset.
10th International Conference on Artificial Intelligence and Applications (AI...gerogepatton
10th International Conference on Artificial Intelligence and Applications (AIFU 2024) is a
forum for presenting new advances and research results in the fields of Artificial Intelligence.
The conference will bring together leading researchers, engineers and scientists in the domain of
interest from around the world. The scope of the conference covers all theoretical and practical
aspects of the Artificial Intelligence.
International Journal of Artificial Intelligence & Applications (IJAIA)gerogepatton
The International Journal of Artificial Intelligence & Applications (IJAIA) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the Artificial Intelligence & Applications (IJAIA). It is an international journal intended for professionals and researchers in all fields of AI for researchers, programmers, and software and hardware manufacturers. The journal also aims to publish new attempts in the form of special issues on emerging areas in Artificial Intelligence and applications.
Hello everyone! I am thrilled to present my latest portfolio on LinkedIn, marking the culmination of my architectural journey thus far. Over the span of five years, I've been fortunate to acquire a wealth of knowledge under the guidance of esteemed professors and industry mentors. From rigorous academic pursuits to practical engagements, each experience has contributed to my growth and refinement as an architecture student. This portfolio not only showcases my projects but also underscores my attention to detail and to innovative architecture as a profession.
Transforming Brand Perception and Boosting Profitabilityaaryangarg12
In today's digital era, the dynamics of brand perception, consumer behavior, and profitability have been profoundly reshaped by the synergy of branding, social media, and website design. This research paper investigates the transformative power of these elements in influencing how individuals perceive brands and products and how this transformation can be harnessed to drive sales and profitability for businesses.
Through an exploration of brand psychology and consumer behavior, this study sheds light on the intricate ways in which effective branding strategies, strategic social media engagement, and user-centric website design contribute to altering consumers' perceptions. We delve into the principles that underlie successful brand transformations, examining how visual identity, messaging, and storytelling can captivate and resonate with target audiences.
Methodologically, this research employs a comprehensive approach, combining qualitative and quantitative analyses. Real-world case studies illustrate the impact of branding, social media campaigns, and website redesigns on consumer perception, sales figures, and profitability. We assess the various metrics, including brand awareness, customer engagement, conversion rates, and revenue growth, to measure the effectiveness of these strategies.
The results underscore the pivotal role of cohesive branding, social media influence, and website usability in shaping positive brand perceptions, influencing consumer decisions, and ultimately bolstering sales and profitability. This paper provides actionable insights and strategic recommendations for businesses seeking to leverage branding, social media, and website design as potent tools to enhance their market position and financial success.
You could be a professional graphic designer and still make mistakes. There is always the possibility of human error. On the other hand if you’re not a designer, the chances of making some common graphic design mistakes are even higher. Because you don’t know what you don’t know. That’s where this blog comes in. To make your job easier and help you create better designs, we have put together a list of common graphic design mistakes that you need to avoid.
The Evolution of Fashion Trends: History to FashionPixel poets
Since time immemorial, fashion has been an ever-evolving form of self-expression, reflecting the cultural, social and economic landscape of different eras. This comprehensive presentation, "The Evolution of Fashion Trends," takes you on a fascinating journey through the dynamic world of fashion trends, exploring how styles have transformed and reinvented themselves over time.
Book Formatting: Quality Control Checks for DesignersConfidence Ago
This presentation was made to help designers who work in publishing houses or format books for printing ensure quality.
Quality control is vital to every industry. This is why every department in a company need create a method they use in ensuring quality. This, perhaps, will not only improve the quality of products and bring errors to the barest minimum, but take it to a near perfect finish.
It is beyond a moot point that a good book will somewhat be judged by its cover, but the content of the book remains king. No matter how beautiful the cover, if the quality of writing or presentation is off, that will be a reason for readers not to come back to the book or recommend it.
So, this presentation points designers to some important things that may be missed by an editor that they could eventually discover and call the attention of the editor.
Expert Accessory Dwelling Unit (ADU) Drafting ServicesResDraft
Whether you’re looking to create a guest house, a rental unit, or a private retreat, our experienced team will design a space that complements your existing home and maximizes your investment. We provide personalized, comprehensive expert accessory dwelling unit (ADU)drafting solutions tailored to your needs, ensuring a seamless process from concept to completion.
The Design Code Google Developer Student Club.pptxadityakushalsaha
This slide is the baby step of UI/UX design. Here you will learn how to research like a designer, how to draw like a designer.
The knowledge can be implemented in anywhere like logo, poster and many more.
Top 5 Indian Style Modular Kitchen DesignsFinzo Kitchens
Get the perfect modular kitchen in Gurgaon at Finzo! We offer high-quality, custom-designed kitchens at the best prices. Wardrobes and home & office furniture are also available. Free consultation! Best Quality Luxury Modular kitchen in Gurgaon available at best price. All types of Modular Kitchens are available U Shaped Modular kitchens, L Shaped Modular Kitchen, G Shaped Modular Kitchens, Inline Modular Kitchens and Italian Modular Kitchen.
Passive Sonar Detection and Classification Based on Demon-Lofar Analysis and Neural Network Algorithms
1. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
DOI:10.5121/ijaia.2024.15106 87
PASSIVE SONAR DETECTION AND CLASSIFICATION
BASED ON DEMON-LOFAR ANALYSIS AND NEURAL
NETWORK ALGORITHMS
Said Jamal, Jawad Lakziz, Yahya Benremdane and Said Ouaskit
Faculty of science Ben M’Sik, University Hassan II, Casablanca, Morocco
ABSTRACT
This paper focuses on an experimental study that used passive sonar sensors as the primary information
source for the submerged target in order to identify, classify, and recognize naval targets. Surface vessels
and submarine generate a specific sound either by propulsion systems, auxiliary equipment or blades of
their propellers, producing information known as the "acoustic signature" that is unique to each type of
target. Consequently, the analysis and classification of targets depend on the processing of the frequencies
produced by these vibrations (sound). utilizing the TPWS (Two-Pass-Split Windows) filter, this work aims
to develop a novel technique for target identification and classification utilizing passive sonars. This
technique involves processing the target's signal in the time-frequency domain. subsequently, in order to
improve the frequency lines of the target noise and decrease the background noise, a TPSW algorithm is
implemented in the frequency domain. By integrating narrowband and broadband analysis as inputs of an
artificial intelligence model that can classify a target into one of the categories given in the training phase,
the target has finally been classified. Our findings demonstrated that the suggested approach is dependent
upon the size of the target noise data collection and the noise-to-effective-signal ratio.
KEYWORDS
Passive sonar, Target analysis, Submerged target, Classification filter, Narrowband Analysis &Artificial
Intelligence.
1. INTRODUCTION
One of the biggest challenges in underwater research is continuing to increase SONAR (SOund
Navigation and Ranging) capabilities. The understanding of underwater life and the advancement
of underwater navigation systems depend on the quick and precise detection and classification of
sonar targets due to the characteristics of acoustic wave propagation and the effects of attenuation
and absorption for sound intensity in the sea [1].
Acoustic waves are a useful information source used in underwater signal processing applications
to locate and identify other vessels as well as access the surrounding environment [2]. Sonar
devices can analyze sound waves in two different ways when it comes to civil or military
maritime navigation: actively and passively [3]. As opposed to the second scenario, which
involves just sound reception rather than sound generation, the first involves the emission of a
multi-frequency pulse, the echo of which is then utilized to identify and categorize potential
targets [4].
Therefore, the passive sonar using only reception of acoustic waves generated by a target is the
only way to classify it, because active sonar gives only distance, bearing and speed of the target.
2. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
88
Sonar systems play a major role in underwater detection. They gather information about the
surrounding environment via sound waves emitted by various types of potential targets with the
ultimate goal of identifying them [5].
The undersea environment produces a lot of noise all the time. Research on underwater acoustic
noise can be conducted in a variety of domains, and each one favors a particular style of
representation based on its requirements and interests. In the context of passive detection, the
examination of underwater noise encompasses not only the features of the waves as studied in
oceanography, but also their investigation as a source of environmental disturbance or
information. The design, implementation, and application of a system for classifying ships and
other naval equipment according to the acoustical signals they emit is also the ultimate objective
[6].
As a result, the sonar operating system's identification and classification accuracy can be
improved and decision-making speed increased with the use of computational techniques. Thus,
the application of novel techniques for the identification of various undersea targets is crucial. In
the realm of machine learning, deep learning is a novel approach that Hinton introduced in 2006
[7]. It has advanced significantly in the last several years in areas like picture recognition and
speech analysis. The capacity of deep learning to extract the deep functionalities concealed in the
target signals using multi-level network architecture without the need for artificially created
structured features is one of its most remarkable aspects. To this objective, the supervised
learning model is employed by Convolution Neural Network (CNN), a well-known deep learning
technique. Y. Lecun's multi-layer learning algorithm CNN has demonstrated success in
handwriting recognition [8]. With remarkable outcomes, the deep-convolution neural network
was applied to Image Net in 2012 [9]. The goal of our effort is to identify the underwater acoustic
targets using this network.
Conversely, the application of LOFAR (LOW Frequency Analysis and Recording) and DEMON
(Detection Envelope Modulation On Noise) analysis [10] has enhanced our comprehension of the
sound retrieved from the SONAR sensor. In this sense, the DEMON analysis sheds light on the
cavitation noise [11] to determine the number of ship blades and the rotation of the propeller
shaft. Furthermore, it is well known that LOFAR analyzes the received sound spectrally, enabling
the presentation of multiple frequency bands at once. So, in the time domain, the LOFARGRAM
matches the LOFAR analysis, allowing for the depiction of spectrum fluctuations over time. It is
possible to display the lines that correspond to the tones in the signal sound and then link them to
the ship's gear.
2. METHODOLOGY
2.1. Theory of the Used Technique
In contemporary digital sonar systems, a single beam performs the processing and categorization
of the target features, whereas several beams are used for signal detection. According to signal
processing theory, the most effective methods for evaluating signals are LOFAR and DEMON
analysis [10].
Sonar must primarily compute three distinct features, beginning with the signal in the temporal
domain:
Broadband signal or continuous spectrum;
LOFAR signal or discrete spectrum;
3. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
89
DEMON signal allows assessing whether there are intermodulation products as cavitation
broadband noise, which is modulated with low frequency lines from the propeller rotation.
A broadband spectral analysis that spans the anticipated range of the controlled object noise is
what LOFAR analysis is. The following is the LOFAR analysis sequence [12]:
Choosing the direction of interest, also called "bearing";
Applying the Hanning window to process the incoming signal;
Applying the Fast Fourier Transform to further process the resultant signal (the FFT
processing is used in order to obtain signal representation in the particular frequency
domain).
Signal normalization is achieved by applying a task-specific method that includes calculating a
normalized frequency interval based on the established normalization factor and assessing the
amount of background noise at each spectrum. Volume and frequency. Peak equalization and the
elimination of signal bias are made possible by this estimation. Conversely, DEMON analysis is
a narrowband analysis that processes cavitation noise while accumulating more data from the
controlled item. In terms of actual use, DEMON analysis makes it possible to determine the
number of shafts, their rotation frequency, and blade rate in addition to separating the cavitation
noise from the entire signal spectrum. This study is also very helpful for target detection since it
offers detailed information about the target propellers. The sequence of DEMON analysis is as
follows:
The steps involved in signal processing are as follows:
Direction of interest selection (also known as "bearing");
Additional band pass filtering to reduce the cavitation frequency range of the overall
signal;
Signal squaring (using a standard demodulation algorithm);
Next, the normalization algorithm is implemented to reduce the background noise level
and highlight the target signal peaks;
Finally, the target signal processing is completed by using a short-time Fast Fourier
Transfer to observe signal peaks in the frequency domain.
2.2. Correction of the Spectrum using an Estimate of the Background Sound
A passive sonar system is typically made from a number of building blocks (see Figure 1);
described in terms of its aim and specific signal processing techniques that have been applied for
signal analysis. [13]
Figure 1. Process of detection and classification in passive sonar
4. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
90
The sound spectrum that is released includes tones that are connected to shipboard gear. These
tones are recorded by SONAR and are placed on a continuous spectrum sound that is created
when nearby background sounds come together. Most of the time, the spectrum is corrected by
estimating the background sound and concentrating on details related to particular amplitude
peaks in the spectrum [14].
Typically, the Two-Pass Split Window (TPSW) technique is used to estimate the background
noise [15]. The average amplitude values are estimated using this approach. First, an estimated
local average is made, with each point denoting the average of its nearest neighbors. The local
average is multiplied by a constant that establishes the detection threshold. The local mean takes
the place of the spectrum points that cross this threshold. A final estimate of the spectrum's
background noise is obtained by performing a second convolution of this updated spectrum with
a fresh window. As previously mentioned, this estimation equalizes the spectrum amplitude and
eliminates spectrum bias.
A way to correct the spectrum by the background sound is specified as follows:
𝒚𝒌
= 𝐥𝐨𝐠 (𝒙𝒌(𝒏)) − 𝑻𝑷𝑺𝑾(𝐥𝐨𝐠 𝒙𝒌
(𝒏))
Where x(n) represents the nth spectrum of class k, and y(n), the corrected spectrum. This
equation permits both correction and normalization of the spectrum.
2.3. Signal Post-Processing
In order to separate narrowband noise from broadband noise, the target signal is treated in the
time-frequency domain in this stage using band-pass filters and decimation blocks. Next, in order
to acquire more suitable results for LOFAR and DEMON viewing, the noise is visualized in a
spectrogram that is augmented using the Two-Pass Split Window technique with precise settings
chosen following laboratory trials.
Lastly, LOFARGRAM and DEMONGRAM are introduced as inputs of a CNN model into a
neural network's training process. This allows the neural network to categorize any signal into
any of the categories that the model has been trained in.
2.4. Experimental Setup
2.4.1. LOFAR Analysis: (Low Frequency Analysis & Recording)
It consists of a broadband analysis to extract propeller and auxiliary machines noises. This
analysis is applied after preprocessing the target noise sound. [16]
Figure 2. LOFAR analysis process
5. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
91
The sonars picked up a signal with a wide frequency range that goes above 40 KHz. Nonetheless,
the [0-1000Hz] band is typically where the relevant frequencies are found. Following the removal
of all other frequencies, we apply Nyquist-Shannon law to decimate the output signal:
𝐅𝐫𝐞𝐪𝒔𝒂𝒎𝒑𝒍𝒊𝒏𝒈 ≥ 𝟐 ∗ 𝑭𝒓𝒆𝒒𝒎𝒂𝒙
After applying a Bartlett window, we divide the signal into blocks of 1024 samples. The Fast
Fourier Transform (FFT) is then used to move the blocks of samples into the frequency domain.
After the spectrum analysis is finished, we compute the energy of each frequency channel [17]
from those values. By comparing those energy values with the threshold, we can isolate the
relevant noise. This brings us to the issue of false alarm detection. Nonetheless, a threshold that
ensures a constant probability of detection and a probability of false detection is desirable for an
even examination of sonar waves throughout all frequency bands. Normalizing the FFT's output
is essential to resolving this problem. The standard deviation in each frequency band from a raw
energy spectrum to a normalized one is calculated to achieve this. Finally, as seen in figure 3, we
are left with a continuous normalized spectrum with distinct peaks.
Figure 3. Form of normalized spectrum
2.4.2. DEMON Analysis: (Detection Envelope Modulation On Noise):
A wideband noise that has been frequency modulated can be found in some noise profiles. Thus,
this analysis is referred to as "Demodulation on noise". These noises are typically associated with
the propeller [18] and the cavitation noise modulation phenomenon produced by the propeller
blade. There is a narrow band [19] of hundreds to thousands of Hz when cavitation noise is
present. The cavitation noise band is then extracted using a band pass filter. After that, the
envelope is extracted using the standard method. The DC component is then taken out. Most of
the time, the sampling rate of the signal is high enough to sample the relevant frequencies at a
higher resolution than is required for observation. Because of this, before using the FFT, the
signal needs to be decimated.
6. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
92
Figure 4. DEMON analysis process
3. RESULTS
The results of the suggested approach are displayed on a database that comprises pre-processing,
neural network conception, learning algorithm selection, network training, and classification, as
shown in Figures 2, 3, and 4.
In this study, we have conducted around one hundred records for each target category (fishing
and merchant vessels) in the environmental noise of the Atlantic sea in order to test and analyze
the effectiveness of the suggested method. The acquired database was utilized to learn the
algorithm. Examples of ambient noise signals in the Atlantic Ocean with a sampling rate of 44
kHz and signals produced by a merchant vessel's propeller are shown in Figures 5 and 6,
respectively.
The architectural model established to achieve our classification is presented in figure 7. This
step consists of a signal processing of the recorded audio based on the two types of spectral
analysis mentioned above, namely: LOFAR analysis and DEMON analysis
To obtain the LOFARGRAM image (Fig. 8), which relates to tonal noise (narrowband noise)
radiated by the target building, and the DEMONGRAM image (Fig. 9), which is likely to take
into account the propeller speed in revolutions/minute as well as the number of blades of the
target.
Figure 5. Sample of ambient noise in Atlantic sea with at 44 Khz
7. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
93
Figure 6. Propeller sound of a merchant vessel
Figure 7. The architecture of Deep Learning used for classification
Figure 8. The LOFARGRAM from frequency 0 to 4200 hz
Figure 9. The DEMONGRAM from frequency 0 to 100 hz
8. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
94
Our proposed classification technique is predicated on a particular kind of Artificial Neural
Network (ANN), which is typically used for image processing and classification. Inspired by the
brain's visual cortex, this kind is known as Network Neural Convolutional.
There are two main sections to the Convolutional Neural Network (CNN), which is a multi-
layered acyclic network (forward feed). An image in the form of a pixel array is supplied as
input. For a grayscale image, it has two dimensions. To portray the primary colors—Red, Green,
and Blue—the color is represented by a third dimension with a depth of three.
The convolutional portion of a CNN is its initial component. It functions as an extractor of image
resources. A series of filters, or kernels, have been applied to a picture to produce new outputs
known as a map of features. Ultimately, the map feature is concatenated and flattened to create a
feature vector. In order to produce a single lengthy feature vector, we flatten the convolutional
layers' output. Additionally, it is linked to the last classification model, sometimes known as an
RNC code or a fully-connected layer. After the convolutional component's output, this CNN code
is coupled to the input of a second part that is made up of completely connected layers (multi-
layered perceptrons).
This section's job is to classify the image by combining the CNN code's features. A final layer
with one neuron for each category is the result. Typically, the derived numerical values are
normalized within the range of 0 to 1, resulting in a probability distribution in categories. Figure
10 depicts the network structure that was used in this article [21].
Figure 10. The architecture of CNN
The next step in CNN training is to reduce the classification error at the output by optimizing the
network's coefficients after random initialization. In actuality, a gradient descent algorithm is
used to modify network coefficients in order to repair the classification errors that are discovered.
The term "back-propagation of the gradient" refers to the process by which these gradients are
used to train neural network algorithms within the network starting at the output layer. It is true
that having a collection of photographs for which we already know which category to choose is
necessary in order to make use of back-propagation. Put otherwise, in order to construct the
LOFAR-GRAM, DEMON-GRAM, and Propeller sounds of each recording, a variety of
recordings of the acoustic noise radiated by the targets of different categories are required.
Having said that, we did manage to get a number of samples from various targets across several
categories.
We initially choose the training data in order to begin learning the RNC. Each sample has three
photos with the appropriate category included. In addition, the maximum number of iterations
that must be completed and the network's maximum learning rate must be specified. Our
application captures the audio of naval targets, processes it, and outputs the LOFARGRAM,
DEMONGRAM, and propeller sound, accordingly. The target can then be classified into one of
9. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
95
the following categories: commerce vessel, warship, or fishing vessel by using the
aforementioned graphs as input vectors. See Figures 11 and 12.
Figure 11. Classification of merchant vessel
Figure 12. Classification of a fishing vessel
In the figures 13, 14 and 15, it is presented the performances in terms of Mean-Square Errors
(MSE), which is the squared difference of the inputs and the outputs of the network. Results
show that the error is more optimized in term of the Iterations Numbers “It”, the momentum
“mom” ,and the number of hiden neurons “nhn” with a step of 0.1.
10. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
96
Figure 13. Mean square error(It=20000, mom=0,997, nhn=4, err=2.07726e-007)
Figure 14. Mean square error(It=80000, mom=0,997, nhn=4, err=1.1802e-025)
Figure 15. Mean square error(It=100000, mom=0,999, nhn=4, err=2.1506e-004)
The deep learning method can achieve significantly better results and a higher accuracy with
93.75% when recognizing the underwater targets by using acoustical recordings.
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
x 10
4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Evolution de l erreur quadratique en fonction du nombre d itérations
0 1 2 3 4 5 6 7 8
x 10
4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Evolution de l erreur quadratique en fonction du nombre d itérations
11. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
97
4. DISCUSSION AND CONCLUSIONS
The primary goal of this study was to create a neural network that could recognize a passive
sonar target's acoustic signature. We took these simple actions in order to do this. Initially, we
divided the signal from the numerous recorded propeller noises into two groups: "fishing vessel"
and "merchant vessel." After that, we developed a program that could produce the propeller
sound, LOFARGRAM, and DEMONGRAM for each category. The images were then saved in
two files with the same names as the categories. In order to train a CNN model to identify the
target's signature, another software is developed. After training was finished, we tested the neural
network scheme by giving it fresh target samples, which enabled us to assess the CNN model's
performance.
The following processes can be observed by a neural network to identify an audio signature,
according to the results. First off, training a neural network can take a very long time. However,
we can shorten this process by optimizing the dataset, modifying the learning parameters, and
selecting the right threshold function. We also discovered that we could greatly improve the CNN
model's overall performance by selecting the optimal neural network design and utilizing the
optimal set of training data. Finally, we observed that neural networks are not infallible. Under
certain circumstances, a target may be mistaken for another target. Retraining the model with an
improved and larger dataset will help to improve this restriction. Additionally, the neural
network's performance can be specified by preparing the input data using the proper signal
processing techniques to increase the signal-to-noise ratio. As a result, neural networks can be
useful in scenarios where human pattern recognition skills are needed; this technology is a good
way to advance sonar systems. To identify various types of ships, the program will incorporate
more vessel categories. Additionally, to enhance passive sonar target classification, several
statistically based signal processing algorithms, like convolutive blind separation [23] and
nonnegative matrix factorization [22], may be used. These elements will be covered in future
project.
REFERENCES
[1] J.Lakziz, S.Otmane, S.Ouaskit, R. El guerjouma, “Finite difference time domain method for
acoustic waves in attenuate and absorptive medium for layered underwater acoustic environments”,
Journal of Marine Technology and Environment 2014, Vol. II, pp. 47-54
[2] R. Urick, Principles of Underwater Sound for Engineers. McGraw-Hill, 1983.
[3] Q. Li, Digital sonar design in underwater acoustics: principles and applications. Springer Science &
Business Media, 2012.
[4] R. O. Nielsen, Sonar Signal Processing. Norwood, MA, USA: Artech House, Inc., 1991.
[5] R. P. Hodges, Underwater acoustics: Analysis, design and performance of sonar. John Wiley &
Sons, 2011.
[6] W. S. Burdic, Underwater Acoustic System Analysis. Peninsula Pub, 2003.
[7] G. E. Hinton, R. R. Salakhutdinov, “Reducing the dimensionality of data with neural networks,”
Science 2006, vol. 313, pp. 504-507.
[8] Y LeCun, B Boser, JS Denker, D Henderson, RE Howard, W Hubbard, LD Jackel ,Handwritten
digit recognition with a back-propagation network, Advances in neural information processing
systems 2, NIPS 1989, pp. 396-404
[9] A.Krizhevsky, I. Sutskever I, G. E. Hinton, “ImageNet classification with deep convolutional neural
networks,” International Conference on Neural Information Processing Systems. Curran Associates
Inc. 2012, pp. 1097-1105.
[10] De Moura, N.N.; De Seixas, J.M.; Ramos, R. Passive Sonar Signal Detection and Classification
Based on Independent Component Analysis. In Sonar Systems; INTECH: Rijeka, Croatia, 2011; pp.
93–104.
12. International Journal of Artificial Intelligence and Applications (IJAIA), Vol.15, No.1, January 2024
98
[11] Propeller cavitation noise investigations of a research vessel using medium size cavitation tunnel
tests and full-scale trials,BatuhanAktasa-Mehmet Atlara-SerkanTurkmena-Weichao-Shi-Roderick
Sampson-EminKorkutPatrick Fitzsimmons, Ocean EngineeringVolume 120, 1 July 2016, Pages
122-135.
[12] http://zetlab.com/en/types-of-sonar-systems-lofar-and-demon-analysis/
[13] N.de Moura, E.S Filho and J.M de Seixas,Independent Component Analysis for Passive Sonar
Signal Processing, Advances in Sonar Technology, (2009) 91-110.
[14] W.Soares,J.M.Seixas,N.Moura,Preprocessing passive sonar signals for neural classification, IET
Radar,Sonar Navig.5 (2011) 605–612.
[15] R.O. Nielsen, Sonar Signal Processing, Artech House, Norwood, MA, USA, 1991.
[16] Di Martino, J.C.; Haton, J.P.; Laporte, A., Lofargram line tracking by multistage decision process.
IEEE International Conference on Acoustics, Speech, and Signal Processing, Minneapolis, USA,
vol. 1, pp. 27-30 April (1993).
[17] Qihu Li, Digital Sonar Design in Underwater Acoustics Principles and Applications.
[18] Nielsen R. O., Sonar Signal Processing, Artech House Inc., Northwood, MA (1991).
[19] BurdicWilliamS., Underwater Acoustics System Analysis, Peninsula Publishing, 2nd Ed. (1991).
[20] RubensL.Oliveiraa,b, BeatrizS.L.P.deLimaa,n, NelsonF.F.Ebecken ,Multiway analysis in data
SONAR classification, Mechanical Systems and Signal Processing45(2014)531–541.
[21] H.Yue, L.Zhang, D.Wang, Y.Wang and Z.Lu. “The classificication of Underwater Acoustic Targets
Based on Deep Learning Methods”, Advanceds in Intelligent Systems Research V.134, 2nd
International Conference on Control, and Artificial Intelligence (CAAI2017), pp.526-529, 2017.
AUTHORS
Mr Said JAMAL Engineer and Phd Student, Faculty of Science Ben M’sik, University
Hassan II, Casablanca, Morocco Said JAMAL is an engineer and Phd student at the
university Hassan II. He’s preparing his thesis on the application of AI in the context of
the classification of submarine acoustic waves.
Dr Jawad LAKZIZ Phd, Faculty of Science Ben M’sik, University Hassan II,
Casablanca, Morocco. Dr Jawad LAKZIZ holds a Phd in Modeling acoustic waves in
submarine environment. He is currently an Associate researcher at the Faculty of Science
Ben M’sik, University Hassan II, Casablanca, Morocco, and guest Editor of the Journal
of Marine Technology and Environment. His research interests are modeling acoustic waves, and signal
processing, including underwater acoustic classification based on IA .
Mr Yahya Benremdane Engineer and Phd Student, Faculty of Science Ben
M’sik,University Hassan II, Casablanca, Morocco As a Telecommunications engineer,
Yahya Benremdane has been working on signal analysis and proprieties of
electromagnetic waves and acoustic waves. He’s interested in new technologies of radio
transmissions, radars, sonars and Electronic Warfare technologies. Since 2019 he’s preparing a thesis on
classification of the modulations of telecommunications based on artificial Intelligence, at the Faculty of
Science Ben M’sik in Casablanca.
Pr Said OUASKIT Professor, Faculty of Science Ben M’sik, University Hassan II,
Casablanca, Morocco Professor Said OUASKIT is a professor at the Faculty of Science
Ben M’sik, University of Hassan II Casablanca. He’s an expert in experimental physics,
condensed Matter, atomic, molecular and Optical Physics.