Pub Date : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549975
Guan Wang, Xianghua Ma
How to make self-driving cars understand the traffic gestures of traffic police is crucial for driverless, especially in China there are many police to help the traffic move smoothly and quickly at different intersection in rush hours. Faster R-CNN in deep learning is a mainstream method, however, has a low recognition rate in the case of complex backgrounds. In order to improve the recognition accuracy under complex environment, a two-stream Faster R-CNN based on color and depth data is proposed in this paper. Depth channel information is used to combine with RGB channel information at the feature level. RGB channel information is integrated with Depth channel information based on Faster R-CNN and RGB-D. Experimental results show that this method is more advantageous than the Faster R-CNN using only RGB data.
{"title":"Traffic Police Gesture Recognition using RGB-D and Faster R-CNN","authors":"Guan Wang, Xianghua Ma","doi":"10.1109/ICIIBMS.2018.8549975","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549975","url":null,"abstract":"How to make self-driving cars understand the traffic gestures of traffic police is crucial for driverless, especially in China there are many police to help the traffic move smoothly and quickly at different intersection in rush hours. Faster R-CNN in deep learning is a mainstream method, however, has a low recognition rate in the case of complex backgrounds. In order to improve the recognition accuracy under complex environment, a two-stream Faster R-CNN based on color and depth data is proposed in this paper. Depth channel information is used to combine with RGB channel information at the feature level. RGB channel information is integrated with Depth channel information based on Faster R-CNN and RGB-D. Experimental results show that this method is more advantageous than the Faster R-CNN using only RGB data.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131248036","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549984
N. Othman, K. Y. Lee, A. Radzol, W. Mansor, U. M. Rashid
Non Structural Protein 1 (NS1) has recently been known as an alternative biomarker for diseases caused by flavivirus. It has been clinically acknowledged for early detection of dengue infection, since NS1 presence in blood can be as early as the first day of infection. Surface Enhanced Raman Spectroscopy (SERS) is an improvement to Raman spectroscopy, which amplifies the intensity of Raman scattering so to be usable. This also enables SERS to detect molecular structure up to a single molecule. As such, it is favorable amongst researchers investigating disease biomarker. Algorithm k-nearest neighbor (kNN) is a strategy to classify an unknown based on learning data, nearest to the class. Our work here intends to determine the optimal nearest neighbor number, distance rule and classifier rule for PCA-EOC-KNN model for automated detection of NS1 fingerprint from SERS spectra of adulterated saliva. Results show that PCA-EOC-KNN classifier performs with accuracy, precision, sensitivity and specificity above 90%, using Consensus classifier rule, Euclidean or Correlation or Cosine distance rule and k-value of 1, 3 and 5.
{"title":"Optimal PCA-EOC-KNN Model for Detection of NS1 from Salivary SERS Spectra","authors":"N. Othman, K. Y. Lee, A. Radzol, W. Mansor, U. M. Rashid","doi":"10.1109/ICIIBMS.2018.8549984","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549984","url":null,"abstract":"Non Structural Protein 1 (NS1) has recently been known as an alternative biomarker for diseases caused by flavivirus. It has been clinically acknowledged for early detection of dengue infection, since NS1 presence in blood can be as early as the first day of infection. Surface Enhanced Raman Spectroscopy (SERS) is an improvement to Raman spectroscopy, which amplifies the intensity of Raman scattering so to be usable. This also enables SERS to detect molecular structure up to a single molecule. As such, it is favorable amongst researchers investigating disease biomarker. Algorithm k-nearest neighbor (kNN) is a strategy to classify an unknown based on learning data, nearest to the class. Our work here intends to determine the optimal nearest neighbor number, distance rule and classifier rule for PCA-EOC-KNN model for automated detection of NS1 fingerprint from SERS spectra of adulterated saliva. Results show that PCA-EOC-KNN classifier performs with accuracy, precision, sensitivity and specificity above 90%, using Consensus classifier rule, Euclidean or Correlation or Cosine distance rule and k-value of 1, 3 and 5.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125326473","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549971
Nwayyin Najat Mohammed, MD. khaleel, M. Latif, Zana Khalid
The principle component analysis(PCA) is a common feature extraction method in machine learning and pattern recognition approaches. PCA has been used in many applications, and face recognition in which specific faces are recognizing in an images database is one of the popular applications. The default distance metric which has been used with PCA based-face recognition is Euclidean distance. In this study, we have tested the Mahalanobis distance instead of Euclidean, and PCA based on Mahalanobis distance suggested a better performance on our students images database with highest recognition rate. However, we proposed weighted and normalized Mahalanobis distance based PCA-face recognition(PCA_WNMD). The proposed algorithm (PCA_WNMD) showed an improvement in faces recognition rate when tested on our students images database compared to PCA based on Mahalanobis and default Euclidean distances.
{"title":"Face Recognition Based on PCA with Weighted and Normalized Mahalanobis distance","authors":"Nwayyin Najat Mohammed, MD. khaleel, M. Latif, Zana Khalid","doi":"10.1109/ICIIBMS.2018.8549971","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549971","url":null,"abstract":"The principle component analysis(PCA) is a common feature extraction method in machine learning and pattern recognition approaches. PCA has been used in many applications, and face recognition in which specific faces are recognizing in an images database is one of the popular applications. The default distance metric which has been used with PCA based-face recognition is Euclidean distance. In this study, we have tested the Mahalanobis distance instead of Euclidean, and PCA based on Mahalanobis distance suggested a better performance on our students images database with highest recognition rate. However, we proposed weighted and normalized Mahalanobis distance based PCA-face recognition(PCA_WNMD). The proposed algorithm (PCA_WNMD) showed an improvement in faces recognition rate when tested on our students images database compared to PCA based on Mahalanobis and default Euclidean distances.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"155 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115995491","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549948
Massimo Marchiori
People flows are of primary importance in a city environment, making up for an essential component of interest in every city. Yet, study of people flows has to face severe problems, mainly due to the high cost/benefit ratio of trying to get flow information. People flows tend to be seen as secondary with respect to traffic in most parts of the cities. The result of this policy is that the detection of their actual status, and corresponding maintenance, is often far from optimal. In this study we tackle the problem of extracting people flow information, and also show a concrete example of usage of the data, that allows to monitor the pedestrian infrastructure of a city. Following the Smart Cheap City (SCC) approach, we design and implement a system of sensors that allows to gather people flow data by staying within a very limited budget. We then show how this raw data can actually be used to reconstruct people flows, and then investigate the relationship between this flow information and the problem of infrastructure monitoring. We experiment with the system in a major experiment involving five cities, using various configurations, and show the effectiveness of the method when used on the field. The overall lesson is that the problem of reconstructing people flows within cities can be faced even when employing very limited resources, also allowing for a better handling of the related transportation infrastructures.
{"title":"People Flow Reconstruction in Cities","authors":"Massimo Marchiori","doi":"10.1109/ICIIBMS.2018.8549948","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549948","url":null,"abstract":"People flows are of primary importance in a city environment, making up for an essential component of interest in every city. Yet, study of people flows has to face severe problems, mainly due to the high cost/benefit ratio of trying to get flow information. People flows tend to be seen as secondary with respect to traffic in most parts of the cities. The result of this policy is that the detection of their actual status, and corresponding maintenance, is often far from optimal. In this study we tackle the problem of extracting people flow information, and also show a concrete example of usage of the data, that allows to monitor the pedestrian infrastructure of a city. Following the Smart Cheap City (SCC) approach, we design and implement a system of sensors that allows to gather people flow data by staying within a very limited budget. We then show how this raw data can actually be used to reconstruct people flows, and then investigate the relationship between this flow information and the problem of infrastructure monitoring. We experiment with the system in a major experiment involving five cities, using various configurations, and show the effectiveness of the method when used on the field. The overall lesson is that the problem of reconstructing people flows within cities can be faced even when employing very limited resources, also allowing for a better handling of the related transportation infrastructures.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123448720","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}
Benign Paroxysmal Positioning Vertigo (BPPV) is one of the causes of vertigo which extremely affects the daily life of patients. Different types of BPPV are treated in a different way. Physicians differentiate the BPPV types using nystagmus characteristics. However, some patients have unclear nystagmus, so their treatments are delayed due to the difficulty of diagnosis. Dizziness Handicap Inventory (DHI) is a tool to assess the severity of dizziness before a patient is diagnosed by a physician. The use of DHI can distinguish BPPV types which can help physicians decide what treatments would be suitable for patients. This research aims to study the ability of using DHI for diferrential diagnosis of Posterior canal — Benign Paroxysmal Positioning Vertigo (PC-BPPV) and Horizontal canal — Benign Paroxysmal Positioning Vertigo (HC-BPPV) via machine learning techniques. We used feature selection techniques and feature engineering to increase the power of machine learning algorithms. Random Forest, Support vector machine, K-Nearest Neighbor and Naïve Bayes were used to develop predictive models from DHI features that have statistically significant. Accuracy, precision, recall, and F1-score were used to evaluate the performance of each model. It reveals that F7+E23, age and P8 are the top three important features and the model using Gaussian Naïve Bayes is the best model to discriminate HC-BPPV and PC-BPPV with 73.91% accuracy, 66.67% precision, 80.00% recall and 72.73% F1-score. In conclusion, the models that were created from DHI score can predict BPPV types at a certain level, but still not very good. Physicians have to use patient�s medical history and nystagmus observation for diagnosis. In the future, if we can collect more data or features, we may reduce the overfitting problem and have a better performance model.
{"title":"Classification of Benign Paroxysmal Positioning Vertigo Types from Dizziness Handicap Inventory using Machine Learning Techniques","authors":"Lawana Masankaran, Waraporn Viyanon, Visan Mahasittiwat","doi":"10.1109/ICIIBMS.2018.8550002","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8550002","url":null,"abstract":"Benign Paroxysmal Positioning Vertigo (BPPV) is one of the causes of vertigo which extremely affects the daily life of patients. Different types of BPPV are treated in a different way. Physicians differentiate the BPPV types using nystagmus characteristics. However, some patients have unclear nystagmus, so their treatments are delayed due to the difficulty of diagnosis. Dizziness Handicap Inventory (DHI) is a tool to assess the severity of dizziness before a patient is diagnosed by a physician. The use of DHI can distinguish BPPV types which can help physicians decide what treatments would be suitable for patients. This research aims to study the ability of using DHI for diferrential diagnosis of Posterior canal — Benign Paroxysmal Positioning Vertigo (PC-BPPV) and Horizontal canal — Benign Paroxysmal Positioning Vertigo (HC-BPPV) via machine learning techniques. We used feature selection techniques and feature engineering to increase the power of machine learning algorithms. Random Forest, Support vector machine, K-Nearest Neighbor and Naïve Bayes were used to develop predictive models from DHI features that have statistically significant. Accuracy, precision, recall, and F1-score were used to evaluate the performance of each model. It reveals that F7+E23, age and P8 are the top three important features and the model using Gaussian Naïve Bayes is the best model to discriminate HC-BPPV and PC-BPPV with 73.91% accuracy, 66.67% precision, 80.00% recall and 72.73% F1-score. In conclusion, the models that were created from DHI score can predict BPPV types at a certain level, but still not very good. Physicians have to use patient�s medical history and nystagmus observation for diagnosis. In the future, if we can collect more data or features, we may reduce the overfitting problem and have a better performance model.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128788246","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549934
T. Oishi, Y. Kuroki
This paper discusses compressed sensing which reconstructs original sparse signal from observed data. Our approach formulates the weighted sum of $l_{1}$ -norm error and $l_{1}$ -norm regularization terms, and applies Alternating Direction Method of Multipliers (ADMM) to solve it. Many works employ ADMM for the $l_{1}-l_{1}$ -norm minimization problems, where ADMM obtains solutions in an iterative fashion for the problems formed as an augmented Lagrangian. The ADMM process is divided into three steps: an error minimization, a coefficient-norm minimization, and a dual variable update of an augmented Lagrangian. However, the coefficient-minimization step is not clear and replaced with an approximation. Our contribution is to adopt the Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) for the minimization step and achieves faster implementation than a conventional method.
{"title":"An $l_{1}-l_{1}$ -Norm Minimization Solution Using ADMM with FISTA","authors":"T. Oishi, Y. Kuroki","doi":"10.1109/ICIIBMS.2018.8549934","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549934","url":null,"abstract":"This paper discusses compressed sensing which reconstructs original sparse signal from observed data. Our approach formulates the weighted sum of $l_{1}$ -norm error and $l_{1}$ -norm regularization terms, and applies Alternating Direction Method of Multipliers (ADMM) to solve it. Many works employ ADMM for the $l_{1}-l_{1}$ -norm minimization problems, where ADMM obtains solutions in an iterative fashion for the problems formed as an augmented Lagrangian. The ADMM process is divided into three steps: an error minimization, a coefficient-norm minimization, and a dual variable update of an augmented Lagrangian. However, the coefficient-minimization step is not clear and replaced with an approximation. Our contribution is to adopt the Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) for the minimization step and achieves faster implementation than a conventional method.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124652377","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8549951
Kefaya Qaddoum
This paper introduces an advanced approach to classify tumor type by microarray gene selection records. The method utilizes gene selection based on shuffling in connection with optimized data clustering. Merging Artificial Bee Colony (ABC) with genetic algorithm (GA) as a clustering tool to choose the key genes develops a new hybrid algorithm, ABC-GA. Support Vector Machine recursive feature elimination (SVM-RFE) and Multilayer Perceptron (MLP) artificial neural networks were used to enhance accuracy. Nonetheless, outcomes show that using shuffling in clustering strengthen classification accuracy significantly. The suggested algorithm (ABC-GA) performs better than Swarm optimization technique (PSO) in reaching good classification results. Better precision has been achieved using (SVM-RFE) classifier against MLP
{"title":"Gene Selection Approach Utilizing Data Clustering Based Technique Optimization for Tumor Classification","authors":"Kefaya Qaddoum","doi":"10.1109/ICIIBMS.2018.8549951","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8549951","url":null,"abstract":"This paper introduces an advanced approach to classify tumor type by microarray gene selection records. The method utilizes gene selection based on shuffling in connection with optimized data clustering. Merging Artificial Bee Colony (ABC) with genetic algorithm (GA) as a clustering tool to choose the key genes develops a new hybrid algorithm, ABC-GA. Support Vector Machine recursive feature elimination (SVM-RFE) and Multilayer Perceptron (MLP) artificial neural networks were used to enhance accuracy. Nonetheless, outcomes show that using shuffling in clustering strengthen classification accuracy significantly. The suggested algorithm (ABC-GA) performs better than Swarm optimization technique (PSO) in reaching good classification results. Better precision has been achieved using (SVM-RFE) classifier against MLP","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117088161","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8588298
Chen Zhou
High Throughput Screening system (HTS) is an emerging class of discrete event systems, usually described by the Max-Plus model. In order to improve system analysis and control efficiency, the Max-Plus model is extended to dioid model , which can efficiently describe and control the system from both time domain and event domain. However, in the actual process of high throughput screening systems with interference, malfunction and other factors, either due to the lack of corresponding sensors, or because it is not possible to directly measure, it may be necessary to try to estimate the state of some systems observation. Therefore, this paper is based on the observer design of high throughput screening system of dioid model to solve such problems. The observer is added to the high-throughput screening system with interference, the optimal observer matrix is obtained according to the residuation theory, and the state estimation of the system with interference is carried out by using the input and output measurements. Finally, an example is given to verify the effectiveness of this observer design method.
{"title":"Observer design of high throughput screening system based on dioid","authors":"Chen Zhou","doi":"10.1109/ICIIBMS.2018.8588298","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8588298","url":null,"abstract":"High Throughput Screening system (HTS) is an emerging class of discrete event systems, usually described by the Max-Plus model. In order to improve system analysis and control efficiency, the Max-Plus model is extended to dioid model , which can efficiently describe and control the system from both time domain and event domain. However, in the actual process of high throughput screening systems with interference, malfunction and other factors, either due to the lack of corresponding sensors, or because it is not possible to directly measure, it may be necessary to try to estimate the state of some systems observation. Therefore, this paper is based on the observer design of high throughput screening system of dioid model to solve such problems. The observer is added to the high-throughput screening system with interference, the optimal observer matrix is obtained according to the residuation theory, and the state estimation of the system with interference is carried out by using the input and output measurements. Finally, an example is given to verify the effectiveness of this observer design method.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115933160","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8550024
May Mon Lynn, C. Su, Kyi Kyi Maw
Recognizing emotion from speech has become the active research themes in speech processing and in applications based on human-computer interaction. The emotion recognition system is also intended to apply in criminal cases; one potential application is the detection of the emotional state in Automatic Teller Machine (ATM) machine to provide feedback to machine. Especially, if it detect fear emotion, it provide feedback to stop all of the money transition. In this domain, the image resolution of facial expressions captured by CCTV camera can be degraded at night due to the inadquate light. To overcome this problem, this paper conducts an experimental study on recognizing emotions from human speech. The emotions considered for the experiments include angry, disgust, happy, fear, sad and surprise. Feature selection method, which aims to extract effective acoustics features to improve the performance of emotion recognition is focused in this paper. The combination of MFCC (Mels Frequency Cepstral Coefficients) with Berouti Spectral Subtraction is used to obtain compressed feature vectors without losing much information. The standard SAVEES dataset and our self-created Myanmar emotional dataset are used for training and testing. After that, those emotions are identified with Random Forest classifier. Finally, the detail analysis results on enhanced features and only MFCCs features corresponding to different databases and different classifiers are compared and explained.
从语音中识别情感已成为语音处理和基于人机交互的应用中活跃的研究课题。该情感识别系统还打算应用于刑事案件;一个潜在的应用是对自动柜员机(ATM)的情绪状态进行检测，并向机器提供反馈。特别是，如果它检测到恐惧情绪，它会提供反馈来阻止所有的金钱转换。在该领域中，由于夜间光线不足，闭路电视摄像机捕捉到的面部表情图像分辨率会降低。为了克服这一问题，本文进行了从人类语言中识别情绪的实验研究。实验中考虑的情绪包括愤怒、厌恶、快乐、恐惧、悲伤和惊讶。本文重点研究了特征选择方法，该方法旨在提取有效的声学特征，以提高情感识别的性能。结合MFCC (Mels Frequency Cepstral Coefficients)和Berouti Spectral Subtraction，在不丢失太多信息的情况下得到压缩后的特征向量。标准的SAVEES数据集和我们自己创建的缅甸情感数据集用于培训和测试。之后，这些情绪被随机森林分类器识别。最后，对不同数据库和不同分类器对应的增强特征和仅mfccc特征的详细分析结果进行了比较和说明。
{"title":"Recognition and Analysis of Emotion Types from Myanmar Movies","authors":"May Mon Lynn, C. Su, Kyi Kyi Maw","doi":"10.1109/ICIIBMS.2018.8550024","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8550024","url":null,"abstract":"Recognizing emotion from speech has become the active research themes in speech processing and in applications based on human-computer interaction. The emotion recognition system is also intended to apply in criminal cases; one potential application is the detection of the emotional state in Automatic Teller Machine (ATM) machine to provide feedback to machine. Especially, if it detect fear emotion, it provide feedback to stop all of the money transition. In this domain, the image resolution of facial expressions captured by CCTV camera can be degraded at night due to the inadquate light. To overcome this problem, this paper conducts an experimental study on recognizing emotions from human speech. The emotions considered for the experiments include angry, disgust, happy, fear, sad and surprise. Feature selection method, which aims to extract effective acoustics features to improve the performance of emotion recognition is focused in this paper. The combination of MFCC (Mels Frequency Cepstral Coefficients) with Berouti Spectral Subtraction is used to obtain compressed feature vectors without losing much information. The standard SAVEES dataset and our self-created Myanmar emotional dataset are used for training and testing. After that, those emotions are identified with Random Forest classifier. Finally, the detail analysis results on enhanced features and only MFCCs features corresponding to different databases and different classifiers are compared and explained.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122271912","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 : 2018-10-01DOI: 10.1109/ICIIBMS.2018.8550008
H. Halin, W. Khairunizam, K. Ikram, Hasri Haris, I. Zunaidi, S. A. Bakar, Z. Razlan, H. Desa
The development of an autonomous vehicle started since 1960 by Stanford University. Steering wheel control for an autonomous vehicle is important for a successful navigation through the designed road/paths. The recent autonomous vehicles developed by Uber, Waymo, Tesla, and others are still in the experimental stage. The current accident with Uber autonomous car in Arizona shows that the safety of the pedestrian and passengers of the autonomous cars are still at risk. The passenger�s comfort and safety while riding in the AEV can be improved by developing a controller that is precise and accurate in making decisions based on the uncertain environment. The developed Fuzzy controller that uses data from the human drivers in order to develop the Fuzzy membership function is one of the possible approaches. The human navigation experiment is the experiment that gathers data from the human driver as they drive through designed paths. The speed, steering wheel angle, heading, and position of the buggy are collected throughout the human navigation experiments. Then, data used to calculate the mean and standard deviation for each membership variables. In order to study the performance of the developed Fuzzy controller, the simulation studies were developed. The simulations are executed by using LabVIEW software. The simulation uses data from human navigation experiments in order to simulate the Fuzzy controller performance. The simulation results are expected to show the minimum path tracking error. Path tracking error is defined as the distance between the vehicle center of gravity and the desired path. The negative path tracking error indicates that the vehicle is to the left of the path. The path tracking error is expected to less than 1 meter.
{"title":"Design Simulation of a Fuzzy Steering Wheel Controller for a buggy car","authors":"H. Halin, W. Khairunizam, K. Ikram, Hasri Haris, I. Zunaidi, S. A. Bakar, Z. Razlan, H. Desa","doi":"10.1109/ICIIBMS.2018.8550008","DOIUrl":"https://doi.org/10.1109/ICIIBMS.2018.8550008","url":null,"abstract":"The development of an autonomous vehicle started since 1960 by Stanford University. Steering wheel control for an autonomous vehicle is important for a successful navigation through the designed road/paths. The recent autonomous vehicles developed by Uber, Waymo, Tesla, and others are still in the experimental stage. The current accident with Uber autonomous car in Arizona shows that the safety of the pedestrian and passengers of the autonomous cars are still at risk. The passenger�s comfort and safety while riding in the AEV can be improved by developing a controller that is precise and accurate in making decisions based on the uncertain environment. The developed Fuzzy controller that uses data from the human drivers in order to develop the Fuzzy membership function is one of the possible approaches. The human navigation experiment is the experiment that gathers data from the human driver as they drive through designed paths. The speed, steering wheel angle, heading, and position of the buggy are collected throughout the human navigation experiments. Then, data used to calculate the mean and standard deviation for each membership variables. In order to study the performance of the developed Fuzzy controller, the simulation studies were developed. The simulations are executed by using LabVIEW software. The simulation uses data from human navigation experiments in order to simulate the Fuzzy controller performance. The simulation results are expected to show the minimum path tracking error. Path tracking error is defined as the distance between the vehicle center of gravity and the desired path. The negative path tracking error indicates that the vehicle is to the left of the path. The path tracking error is expected to less than 1 meter.","PeriodicalId":430326,"journal":{"name":"2018 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS)","volume":"276 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124447970","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：