Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454200
A. Rao, H. Fishman
Identifying diseases in Optical Coherence Tomography (OCT) images using Deep Learning models and methods is emerging as a powerful technique to enhance clinical diagnosis. Identifying macular diseases in the eye at an early stage and preventing misdiagnosis is crucial. The current methods developed for OCT image analysis have not yet been integrated into an accessible form-factor that can be utilized in a real-life scenario by Ophthalmologists. Additionally, current methods do not employ robust multiple metric feedback. This paper proposes a highly accurate smartphone-based Deep Learning system, OCTAI, that allows a user to take an OCT picture and receive real-time feedback through on-device inference. OCTAI analyzes the input OCT image in three different ways: (1) full image analysis, (2) quadrant based analysis, and (3) disease detection based analysis. With these three analysis methods, along with an Ophthalmologist's interpretation, a robust diagnosis can potentially be made. The ultimate goal of OCTAI is to assist Ophthalmologists in making a diagnosis through a digital second opinion and enabling them to cross-check their diagnosis before making a decision based on purely manual analysis of OCT images. OCTAI has the potential to allow Ophthalmologists to improve their diagnosis and may reduce misdiagnosis rates, leading to faster treatment of diseases.
{"title":"OCTAI: Smartphone-based Optical Coherence Tomography Image Analysis System","authors":"A. Rao, H. Fishman","doi":"10.1109/AIIoT52608.2021.9454200","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454200","url":null,"abstract":"Identifying diseases in Optical Coherence Tomography (OCT) images using Deep Learning models and methods is emerging as a powerful technique to enhance clinical diagnosis. Identifying macular diseases in the eye at an early stage and preventing misdiagnosis is crucial. The current methods developed for OCT image analysis have not yet been integrated into an accessible form-factor that can be utilized in a real-life scenario by Ophthalmologists. Additionally, current methods do not employ robust multiple metric feedback. This paper proposes a highly accurate smartphone-based Deep Learning system, OCTAI, that allows a user to take an OCT picture and receive real-time feedback through on-device inference. OCTAI analyzes the input OCT image in three different ways: (1) full image analysis, (2) quadrant based analysis, and (3) disease detection based analysis. With these three analysis methods, along with an Ophthalmologist's interpretation, a robust diagnosis can potentially be made. The ultimate goal of OCTAI is to assist Ophthalmologists in making a diagnosis through a digital second opinion and enabling them to cross-check their diagnosis before making a decision based on purely manual analysis of OCT images. OCTAI has the potential to allow Ophthalmologists to improve their diagnosis and may reduce misdiagnosis rates, leading to faster treatment of diseases.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134463102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454235
Kyle Ezekiel S. Juadines, A. Ballado, E. Macalalad, Merlin M. Mendoza
In this paper, a web-based application for calculating ionospheric scintillation proxy indexes (S4p-l and S4p-2) and rate of total electron content index (RoTI) was developed using MATLAB. Ionospheric scintillation is a phenomenon described by rapid temporal fluctuations of incoming radio wave signals passing through irregularities in the ionosphere. Through the internet, this application can accept multiple high-rate GNSS data (compressed RINEX file) maximum of 1-day data (96 GNSS files), extracting the pseudo-range numbers, carrier-to-noise ratio (C/No) data, phase measurement data, and accept GPS daily navigational data (brdc) and then calculate and graph the ionospheric scintillation proxy indexes and RoTI.
{"title":"Development of a MATLAB Web-based Application for Calculating Ionospheric Scintillation Proxy Indexes (S4p-1 and S4p-2) and the Rate of Total Electron Content Index (RoTI)","authors":"Kyle Ezekiel S. Juadines, A. Ballado, E. Macalalad, Merlin M. Mendoza","doi":"10.1109/AIIoT52608.2021.9454235","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454235","url":null,"abstract":"In this paper, a web-based application for calculating ionospheric scintillation proxy indexes (S4p-l and S4p-2) and rate of total electron content index (RoTI) was developed using MATLAB. Ionospheric scintillation is a phenomenon described by rapid temporal fluctuations of incoming radio wave signals passing through irregularities in the ionosphere. Through the internet, this application can accept multiple high-rate GNSS data (compressed RINEX file) maximum of 1-day data (96 GNSS files), extracting the pseudo-range numbers, carrier-to-noise ratio (C/No) data, phase measurement data, and accept GPS daily navigational data (brdc) and then calculate and graph the ionospheric scintillation proxy indexes and RoTI.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129335574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454248
N. Tasneem, Md Anik Hasan, Sumaiya Binte Akther, Mohammad Monirujjaman Khan
The work aims to propose a research and development project to prepare a device that can test sample of soil from the field and the testing sample will give suggestion about required amount of fertilizer. The device will work as a datasheet for the users. In the soil sample, the proposed device will measure the level of different nutrients in the soil NPK (nitrogen, phosphorus and potassium) values using the Beer's law method. An Arduino is used for microcontroller. Different light colors are used to light up watery soil solution under testing. Light gets bounce back from solution. It depends upon its absorbent reflectance of soil. Reflected light is received by another Light Depending Resistor (LDR) which is converted into electrical signal. With the help of Microcontroller, the proposed device can measure soil nutrients. It will be helpful for the farmers to cultivate in a smart way to get more quality products.
{"title":"An Automatic Soil Testing Machine for Accurate Fertilization","authors":"N. Tasneem, Md Anik Hasan, Sumaiya Binte Akther, Mohammad Monirujjaman Khan","doi":"10.1109/AIIoT52608.2021.9454248","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454248","url":null,"abstract":"The work aims to propose a research and development project to prepare a device that can test sample of soil from the field and the testing sample will give suggestion about required amount of fertilizer. The device will work as a datasheet for the users. In the soil sample, the proposed device will measure the level of different nutrients in the soil NPK (nitrogen, phosphorus and potassium) values using the Beer's law method. An Arduino is used for microcontroller. Different light colors are used to light up watery soil solution under testing. Light gets bounce back from solution. It depends upon its absorbent reflectance of soil. Reflected light is received by another Light Depending Resistor (LDR) which is converted into electrical signal. With the help of Microcontroller, the proposed device can measure soil nutrients. It will be helpful for the farmers to cultivate in a smart way to get more quality products.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127421080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454234
Shaheer Ansari, A. Ayob, M. Lipu, M. Saad, A. Hussain
The Remaining Useful Life (RUL) of a battery is very important factor to allow for efficient working of all associated systems. In this paper, a Multi-Battery Input Profile (MBIP) based Cascade Forward Neural Network (CFNN) model is proposed to predict the RUL of Lithium-ion battery. The proposed model was trained by utilizing the NASA battery datasets. In addition, systematic sampling was observed to extract the data from the parameters of charging profile of the battery. Four batteries namely B0005, B0006, B0007 and B0018 are utilized and experiment was performed while training the model with 70:30 ratios. The prediction accuracy of the model in case of B0006 and B0018 was lower as compared with B0005 and B0007 due to the effect of capacity regeneration phenomena. The proposed methodology of CFNN based MBIP is validated with Single-Battery Input Profile (SBIP). Several performance metrics such as Root Mean Square Error (RMSE), Mean Squared Error (MSE) and Mean Absolute Error (MAE) are observed.
{"title":"A Comparative Analysis of Lithium Ion Battery Input Profiles for Remaining Useful Life Prediction by Cascade Forward Neural Network","authors":"Shaheer Ansari, A. Ayob, M. Lipu, M. Saad, A. Hussain","doi":"10.1109/AIIoT52608.2021.9454234","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454234","url":null,"abstract":"The Remaining Useful Life (RUL) of a battery is very important factor to allow for efficient working of all associated systems. In this paper, a Multi-Battery Input Profile (MBIP) based Cascade Forward Neural Network (CFNN) model is proposed to predict the RUL of Lithium-ion battery. The proposed model was trained by utilizing the NASA battery datasets. In addition, systematic sampling was observed to extract the data from the parameters of charging profile of the battery. Four batteries namely B0005, B0006, B0007 and B0018 are utilized and experiment was performed while training the model with 70:30 ratios. The prediction accuracy of the model in case of B0006 and B0018 was lower as compared with B0005 and B0007 due to the effect of capacity regeneration phenomena. The proposed methodology of CFNN based MBIP is validated with Single-Battery Input Profile (SBIP). Several performance metrics such as Root Mean Square Error (RMSE), Mean Squared Error (MSE) and Mean Absolute Error (MAE) are observed.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129773738","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454214
Huda Ali Alatwi, A. Aldweesh
Due to their massive success in various domains, deep learning techniques are increasingly used to design network intrusion detection solutions that detect and mitigate unknown and known attacks with high accuracy detection rates and minimal feature engineering. However, it has been found that deep learning models are vulnerable to data instances that can mislead the model to make incorrect classification decisions socalled adversarial examples. Such vulnerability allows attackers to target NIDSs in a black-box setting by adding small crafty perturbations to the malicious traffic to evade detection and disrupt the system's critical functionalities. Yet, little researches have addressed the risks of black-box adversarial attacks against NIDS and proposed mitigation solutions. This survey explores this research problem and identifies open issues and certain areas that demand further research for considerable impacts.
{"title":"Adversarial Black-Box Attacks Against Network Intrusion Detection Systems: A Survey","authors":"Huda Ali Alatwi, A. Aldweesh","doi":"10.1109/AIIoT52608.2021.9454214","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454214","url":null,"abstract":"Due to their massive success in various domains, deep learning techniques are increasingly used to design network intrusion detection solutions that detect and mitigate unknown and known attacks with high accuracy detection rates and minimal feature engineering. However, it has been found that deep learning models are vulnerable to data instances that can mislead the model to make incorrect classification decisions socalled adversarial examples. Such vulnerability allows attackers to target NIDSs in a black-box setting by adding small crafty perturbations to the malicious traffic to evade detection and disrupt the system's critical functionalities. Yet, little researches have addressed the risks of black-box adversarial attacks against NIDS and proposed mitigation solutions. This survey explores this research problem and identifies open issues and certain areas that demand further research for considerable impacts.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128337248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454203
Shadan Ghaffaripour, A. Miri
Reliability is a crucial component to machine-learning-as-a-service platforms, as more and more critical applications depend on them. Thus, mechanisms employed to assure the integrity of computations performed on such platforms are pivotal to their robust functioning. Moreover, privacy protection, and performance guarantee at scale, are other major challenges surrounding these platforms that are by no means straightforward to overcome at the same time. In this paper, we have proposed a novel distributed approach, which uses specialized composable proof systems at its core, to respond to these challenges. At a high level, we adopt a divide-and-conquer approach to build efficient proof systems for machine-learning-based services in order to ensure the correctness of results. More precisely, the mathematical formulation of the machine learning task is divided into multiple parts, each of which is handled by a different specialized proof system; these proof systems are then combined with the commit-and-prove methodology to guarantee correctness as a whole. With privacy safeguards built into the design, our approach also assures that neither user data nor model parameters, which constitute the intellectual property of service providers are exposed in the process. We have showcased the usability of our approach within a machine learning service provider that offers classification services through a linear support vector machine (SVM) model. Our complexity analysis indicates that our system could be used in practical settings.
{"title":"Mutually Private Verifiable Machine Learning As-a-service: A Distributed Approach","authors":"Shadan Ghaffaripour, A. Miri","doi":"10.1109/AIIoT52608.2021.9454203","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454203","url":null,"abstract":"Reliability is a crucial component to machine-learning-as-a-service platforms, as more and more critical applications depend on them. Thus, mechanisms employed to assure the integrity of computations performed on such platforms are pivotal to their robust functioning. Moreover, privacy protection, and performance guarantee at scale, are other major challenges surrounding these platforms that are by no means straightforward to overcome at the same time. In this paper, we have proposed a novel distributed approach, which uses specialized composable proof systems at its core, to respond to these challenges. At a high level, we adopt a divide-and-conquer approach to build efficient proof systems for machine-learning-based services in order to ensure the correctness of results. More precisely, the mathematical formulation of the machine learning task is divided into multiple parts, each of which is handled by a different specialized proof system; these proof systems are then combined with the commit-and-prove methodology to guarantee correctness as a whole. With privacy safeguards built into the design, our approach also assures that neither user data nor model parameters, which constitute the intellectual property of service providers are exposed in the process. We have showcased the usability of our approach within a machine learning service provider that offers classification services through a linear support vector machine (SVM) model. Our complexity analysis indicates that our system could be used in practical settings.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131162738","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454198
H. Qusa, Jumana Tarazi
The increased digitization of several critical infrastructure services on Internet like home banking, online payments, etc. exposes them to a range of sophisticated information security attacks. Thus, there is an urgent need for strong collaboration between all governmental and non-governmental organizations in order to form the defenses by sharing. Sharing and analyzing sensitive traffic data is an important aspect to protect critical infrastructures. However, privacy concerns of the data contributors about sharing their sensitive data prevent them from gaining the benefits from collaboration, or at least weaken it to a degree of insufficiency. To cope with those privacy concerns, we extend our preceding work about constructing an efficient framework for personal collaborative event processing permitting information sharing and processing amongst administratively and geographically disjoint organizations. The structure is able to aggregating and correlating events coming from the organizations in near real-time while preserving the privacy of sensitive data even in case of coalition among the entities in the environment. The key novelty of the structure is the use of a pseudorandom oracle capability dispensed among the use of FOG structure among the organizations collaborating to the system for obfuscating the data, that permits for achieving a good level of privacy at the same time as guaranteeing scalability in both dimensions: horizontally (range of collaborators) and vertically (range of dataset per collaborator).
{"title":"Collaborative Fog Computing Architecture for Privacy-Preserving Data Aggregation","authors":"H. Qusa, Jumana Tarazi","doi":"10.1109/AIIoT52608.2021.9454198","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454198","url":null,"abstract":"The increased digitization of several critical infrastructure services on Internet like home banking, online payments, etc. exposes them to a range of sophisticated information security attacks. Thus, there is an urgent need for strong collaboration between all governmental and non-governmental organizations in order to form the defenses by sharing. Sharing and analyzing sensitive traffic data is an important aspect to protect critical infrastructures. However, privacy concerns of the data contributors about sharing their sensitive data prevent them from gaining the benefits from collaboration, or at least weaken it to a degree of insufficiency. To cope with those privacy concerns, we extend our preceding work about constructing an efficient framework for personal collaborative event processing permitting information sharing and processing amongst administratively and geographically disjoint organizations. The structure is able to aggregating and correlating events coming from the organizations in near real-time while preserving the privacy of sensitive data even in case of coalition among the entities in the environment. The key novelty of the structure is the use of a pseudorandom oracle capability dispensed among the use of FOG structure among the organizations collaborating to the system for obfuscating the data, that permits for achieving a good level of privacy at the same time as guaranteeing scalability in both dimensions: horizontally (range of collaborators) and vertically (range of dataset per collaborator).","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"34 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120807442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454218
M. Arozullah, Hong-Fang Yu
The wireless network is motivated by the rapidly increasing demand for 5G network. The 5G wireless network system is expect to support different resource services sorted Constant Bit Rate, Variable Bit Rate, Available Bit Rate and Unspecific Bit Rate Request of Services. In this paper, we propose a multiple access protocol for the signal with an integrated mix of multimedia traffic in the 5G wireless network. When the buffer of base station is empty in high speed network system, services are assigned priority class on the base of Time-to-Live (TTL) with respect to the service source types within each TTL class. There is a need of piggybacking request with the package transmissions synchronously when the packets arrive at a non-empty buffer. It also shows the transmission requests are placed collision-free. The multiple access schemes are very challenging such as more efficiency, intelligent, no subsequence collision and cybersecurity. The expect simulation results are evaluate the packet throughput, packet loss and packet delay. The results illustrate the proposed scheme has better performance than the conventional packet reservation multiple access, distributed queuing request update multiple access and adaptive request channel multiple access for the 5G wireless network.
{"title":"A Multiple Access Protocol for Multimedia Transmission over 5G Wireless Asynchronous Transfer Mode Network","authors":"M. Arozullah, Hong-Fang Yu","doi":"10.1109/AIIoT52608.2021.9454218","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454218","url":null,"abstract":"The wireless network is motivated by the rapidly increasing demand for 5G network. The 5G wireless network system is expect to support different resource services sorted Constant Bit Rate, Variable Bit Rate, Available Bit Rate and Unspecific Bit Rate Request of Services. In this paper, we propose a multiple access protocol for the signal with an integrated mix of multimedia traffic in the 5G wireless network. When the buffer of base station is empty in high speed network system, services are assigned priority class on the base of Time-to-Live (TTL) with respect to the service source types within each TTL class. There is a need of piggybacking request with the package transmissions synchronously when the packets arrive at a non-empty buffer. It also shows the transmission requests are placed collision-free. The multiple access schemes are very challenging such as more efficiency, intelligent, no subsequence collision and cybersecurity. The expect simulation results are evaluate the packet throughput, packet loss and packet delay. The results illustrate the proposed scheme has better performance than the conventional packet reservation multiple access, distributed queuing request update multiple access and adaptive request channel multiple access for the 5G wireless network.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122119162","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454174
Damian Valles, Rezwan Matin
Children with Autism Spectrum Disorder (ASD) find it difficult to detect human emotions in social interactions. A speech emotion recognition system was developed in this work, which aims to help these children to better identify the emotions of their communication partner. The system was developed using machine learning and deep learning techniques. Through the use of ensemble learning, multiple machine learning algorithms were joined to provide a final prediction on the recorded input utterances. The ensemble of models includes a Support Vector Machine (SVM), a Multi-Layer Perceptron (MLP), and a Recurrent Neural Network (RNN). All three models were trained on the Ryerson Audio-Visual Database of Emotional Speech and Songs (RAVDESS), the Toronto Emotional Speech Set (TESS), and the Crowd-sourced Emotional Multimodal Actors Dataset (CREMA-D). A fourth dataset was used, which was created by adding background noise to the clean speech files from the datasets previously mentioned. The paper describes the audio processing of the samples, the techniques used to include the background noise, and the feature extraction coefficients considered for the development and training of the models. This study presents the performance evaluation of the individual models to each of the datasets, inclusion of the background noises, and the combination of using all of the samples in all three datasets. The evaluation was made to select optimal hyperparameters configuration of the models to evaluate the performance of the ensemble learning approach through majority voting. The overall performance of the ensemble learning reached a peak accuracy of 66.5%, reaching a higher performance emotion classification accuracy than the MLP model which reached 65.7%.
{"title":"An Audio Processing Approach using Ensemble Learning for Speech-Emotion Recognition for Children with ASD","authors":"Damian Valles, Rezwan Matin","doi":"10.1109/AIIoT52608.2021.9454174","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454174","url":null,"abstract":"Children with Autism Spectrum Disorder (ASD) find it difficult to detect human emotions in social interactions. A speech emotion recognition system was developed in this work, which aims to help these children to better identify the emotions of their communication partner. The system was developed using machine learning and deep learning techniques. Through the use of ensemble learning, multiple machine learning algorithms were joined to provide a final prediction on the recorded input utterances. The ensemble of models includes a Support Vector Machine (SVM), a Multi-Layer Perceptron (MLP), and a Recurrent Neural Network (RNN). All three models were trained on the Ryerson Audio-Visual Database of Emotional Speech and Songs (RAVDESS), the Toronto Emotional Speech Set (TESS), and the Crowd-sourced Emotional Multimodal Actors Dataset (CREMA-D). A fourth dataset was used, which was created by adding background noise to the clean speech files from the datasets previously mentioned. The paper describes the audio processing of the samples, the techniques used to include the background noise, and the feature extraction coefficients considered for the development and training of the models. This study presents the performance evaluation of the individual models to each of the datasets, inclusion of the background noises, and the combination of using all of the samples in all three datasets. The evaluation was made to select optimal hyperparameters configuration of the models to evaluate the performance of the ensemble learning approach through majority voting. The overall performance of the ensemble learning reached a peak accuracy of 66.5%, reaching a higher performance emotion classification accuracy than the MLP model which reached 65.7%.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117152820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-05-10DOI: 10.1109/AIIoT52608.2021.9454168
Khalid Albulayhi, Frederick T. Sheldon
Nowadays, IoT technology has become an essential part of many aspects of life and business. Nevertheless, such widespread application has come at the cost of many security concerns that threaten data privacy and diminish IoT utilization momentum in critical applications such as the smart grid and intelligent transportation systems. To address this challenge, several approaches have been proposed to detect and prevent IoT cyberthreats from materializing. Anomaly detection is one of these approaches that defines the boundaries of legitimate (normal) behavior. Any behavior that falls outside these boundaries is considered anomalous. However, these solutions should have the capability to adapt and adjust to environmental changes that prompt IoT nodal behavioral aberrations, except they only assume that these nodes show the same behavior. This assumption does not hold due to the heterogeneity of IoT nodes and the dynamic nature of an IoT network topology. Furthermore, existing adaptive solutions rely on static (pre-defined) thresholds to control the moment for retraining updates. The cost is heavy for highly dynamic environments like IoT as it leads to an unnecessary higher frequency of retraining. Consequently, the model becomes unstable and adversely affects its accuracy and robustness. This paper addresses these problems by offering an improved Adaptive Anomaly Detection (AAD) methodology that resolves the heterogeneity issues by building local profiles that define normal behavior at each IoT node. The One Class Support Vector Machines (OC-SVM) was used to build these profiles. Then, K-Means clustering was used to build a global profile that represents all network nodes. A Local-Global Ratio-Based (LGR) Anomaly Detection scheme is advanced and was enlisted to control the adaptation process by adjusting the threshold of adaptive functionality dynamically based on the “current” situation to prevent unnecessary retraining. An Ensemble of Deep Belief Networks (EDBN) is developed and used to train the anomaly detection model. Additionally, this study's proposes a new Minimized Redundancy Discriminative Feature Selection (MRD-FS) technique to resolve the issue of redundant features. The MRD-FS experimental evaluation shows detection accuracy higher than those of the related solutions including lower false alarm rates. This validates the efficacy of the proposed model for various IoT applications such as smart grids, smart homes, smart cities and intelligent transportation systems.
{"title":"An Adaptive Deep-Ensemble Anomaly-Based Intrusion Detection System for the Internet of Things","authors":"Khalid Albulayhi, Frederick T. Sheldon","doi":"10.1109/AIIoT52608.2021.9454168","DOIUrl":"https://doi.org/10.1109/AIIoT52608.2021.9454168","url":null,"abstract":"Nowadays, IoT technology has become an essential part of many aspects of life and business. Nevertheless, such widespread application has come at the cost of many security concerns that threaten data privacy and diminish IoT utilization momentum in critical applications such as the smart grid and intelligent transportation systems. To address this challenge, several approaches have been proposed to detect and prevent IoT cyberthreats from materializing. Anomaly detection is one of these approaches that defines the boundaries of legitimate (normal) behavior. Any behavior that falls outside these boundaries is considered anomalous. However, these solutions should have the capability to adapt and adjust to environmental changes that prompt IoT nodal behavioral aberrations, except they only assume that these nodes show the same behavior. This assumption does not hold due to the heterogeneity of IoT nodes and the dynamic nature of an IoT network topology. Furthermore, existing adaptive solutions rely on static (pre-defined) thresholds to control the moment for retraining updates. The cost is heavy for highly dynamic environments like IoT as it leads to an unnecessary higher frequency of retraining. Consequently, the model becomes unstable and adversely affects its accuracy and robustness. This paper addresses these problems by offering an improved Adaptive Anomaly Detection (AAD) methodology that resolves the heterogeneity issues by building local profiles that define normal behavior at each IoT node. The One Class Support Vector Machines (OC-SVM) was used to build these profiles. Then, K-Means clustering was used to build a global profile that represents all network nodes. A Local-Global Ratio-Based (LGR) Anomaly Detection scheme is advanced and was enlisted to control the adaptation process by adjusting the threshold of adaptive functionality dynamically based on the “current” situation to prevent unnecessary retraining. An Ensemble of Deep Belief Networks (EDBN) is developed and used to train the anomaly detection model. Additionally, this study's proposes a new Minimized Redundancy Discriminative Feature Selection (MRD-FS) technique to resolve the issue of redundant features. The MRD-FS experimental evaluation shows detection accuracy higher than those of the related solutions including lower false alarm rates. This validates the efficacy of the proposed model for various IoT applications such as smart grids, smart homes, smart cities and intelligent transportation systems.","PeriodicalId":443405,"journal":{"name":"2021 IEEE World AI IoT Congress (AIIoT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128080879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：