Pub Date : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786601
M. Dehghani, M. Mardaneh, O. Malik, S. M. NouraeiPour
Metaheuristic optimization algorithms have been used in many applications in recent years. Most of these algorithms are inspired by physical processes or living beings' behaviors. A new optimization algorithm, called Donkey Theorem Optimization (DTO), that simulates the behavior of Donkeys is proposed in this paper. DTO is based on donkey theorem that mimics behavior of donkey for reach to food. Proposed algorithm is tested on 23 well-known benchmark test functions and its performance compared with eight optimization algorithms. The results show that DTO is able to provide better results as compared to the other well-known optimization algorithms.
{"title":"DTO: Donkey Theorem Optimization","authors":"M. Dehghani, M. Mardaneh, O. Malik, S. M. NouraeiPour","doi":"10.1109/IranianCEE.2019.8786601","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786601","url":null,"abstract":"Metaheuristic optimization algorithms have been used in many applications in recent years. Most of these algorithms are inspired by physical processes or living beings' behaviors. A new optimization algorithm, called Donkey Theorem Optimization (DTO), that simulates the behavior of Donkeys is proposed in this paper. DTO is based on donkey theorem that mimics behavior of donkey for reach to food. Proposed algorithm is tested on 23 well-known benchmark test functions and its performance compared with eight optimization algorithms. The results show that DTO is able to provide better results as compared to the other well-known optimization algorithms.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"26 1","pages":"1855-1859"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78541923","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786720
Behnam Bakhshi, H. Veisi
Fingerprint recognition has become one of the most reliable ways for human identification due to its uniqueness and consistency. The fingerprint matching problem is formulated as a classification system in which a model is learned to classify every two fingerprints as a genuine or impostor pair. Traditional approaches perform a feature extraction step before matching a fingerprint pair. On the other hand, recently convolutional neural networks (CNNs) have presented exceptional success for many image processing tasks such as face recognition. However, there have been only a few attempts to develop fully CNN methods to deal with challenges in fingerprint recognition problem. In this paper, a CNN-based fingerprint matching method has been developed. A key contribution of the proposed method is to directly learn fingerprint patterns from raw pixels of images. In order to achieve robustness and characterize the similarities comprehensively, incomplete and partial fingerprint pairs were taken into account to extract complementary features. Also, we proposed an end to end CNN approach that contains the feature extraction part of the trained AlexNet network. The network reached an EER of 17.5% on the FVC2002 dataset, that shows better results in comparison to the MinutiaSC and A-KAZE methods.
{"title":"End to End Fingerprint Verification Based on Convolutional Neural Network","authors":"Behnam Bakhshi, H. Veisi","doi":"10.1109/IranianCEE.2019.8786720","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786720","url":null,"abstract":"Fingerprint recognition has become one of the most reliable ways for human identification due to its uniqueness and consistency. The fingerprint matching problem is formulated as a classification system in which a model is learned to classify every two fingerprints as a genuine or impostor pair. Traditional approaches perform a feature extraction step before matching a fingerprint pair. On the other hand, recently convolutional neural networks (CNNs) have presented exceptional success for many image processing tasks such as face recognition. However, there have been only a few attempts to develop fully CNN methods to deal with challenges in fingerprint recognition problem. In this paper, a CNN-based fingerprint matching method has been developed. A key contribution of the proposed method is to directly learn fingerprint patterns from raw pixels of images. In order to achieve robustness and characterize the similarities comprehensively, incomplete and partial fingerprint pairs were taken into account to extract complementary features. Also, we proposed an end to end CNN approach that contains the feature extraction part of the trained AlexNet network. The network reached an EER of 17.5% on the FVC2002 dataset, that shows better results in comparison to the MinutiaSC and A-KAZE methods.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"23 1","pages":"1994-1998"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76759631","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786667
Saba Bayatfar, Saman Seifpour, M. A. Oskoei, Ali Khadem
Today, a large number of people all around the world are affected or threatened by different kinds of sleep disorders, especially sleep apnea syndrome (SAS). Since EEG activities during sleep exhibit physiological and pathological situation of the brain, analyzing the oscillations during various sleep stages are of fundamental importance for the assessment of sleep disorders. On the other hand, recent developments in the fields of signal processing and pattern recognition have prompted sleep researchers to employ accurate and reliable computerized systems in diagnosis processes. The purpose of this paper was to develop a novel automated methodology for diagnosing apneic patients from healthy subjects using single-channel EEG signal. To this end, it was concentrated on the rapid eye movement (REM) sleep. A comprehensive time-domain based feature extraction scheme was proposed for extracting features from various EEG frequency bands of REM sleep. Uninformative features were eliminated by minimal redundancy maximal relevance (mRMR) feature selection algorithm. An ensemble learning method namely random undersampling boosting (RUSBoost) was then utilized for the purpose of classification. The effectiveness and generalizability of the proposed approach were evaluated using k-fold cross-validation technique. The experimental results demonstrate that the proposed scheme can be used as an efficient method for automated SAS screening with regard to promising classification performance in terms of accuracy, sensitivity, and specificity.
{"title":"An Automated System for Diagnosis of Sleep Apnea Syndrome Using Single-Channel EEG Signal","authors":"Saba Bayatfar, Saman Seifpour, M. A. Oskoei, Ali Khadem","doi":"10.1109/IranianCEE.2019.8786667","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786667","url":null,"abstract":"Today, a large number of people all around the world are affected or threatened by different kinds of sleep disorders, especially sleep apnea syndrome (SAS). Since EEG activities during sleep exhibit physiological and pathological situation of the brain, analyzing the oscillations during various sleep stages are of fundamental importance for the assessment of sleep disorders. On the other hand, recent developments in the fields of signal processing and pattern recognition have prompted sleep researchers to employ accurate and reliable computerized systems in diagnosis processes. The purpose of this paper was to develop a novel automated methodology for diagnosing apneic patients from healthy subjects using single-channel EEG signal. To this end, it was concentrated on the rapid eye movement (REM) sleep. A comprehensive time-domain based feature extraction scheme was proposed for extracting features from various EEG frequency bands of REM sleep. Uninformative features were eliminated by minimal redundancy maximal relevance (mRMR) feature selection algorithm. An ensemble learning method namely random undersampling boosting (RUSBoost) was then utilized for the purpose of classification. The effectiveness and generalizability of the proposed approach were evaluated using k-fold cross-validation technique. The experimental results demonstrate that the proposed scheme can be used as an efficient method for automated SAS screening with regard to promising classification performance in terms of accuracy, sensitivity, and specificity.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"6 1","pages":"1829-1833"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82092916","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786763
Mahdi Chegnizadeh, M. Memarian, K. Mehrany
A general and closed form solution of wave propagation in a metallic slab waveguide undergoing any arbitrary, time-varying function of permittivity is presented. The formulation is general and accurate to account for all temporal variations of permittivity. Parametric amplification and frequency conversion is demonstrated using this closed form formulation. Moreover, all results are validated and confirmed using full wave FDTD simulation. The in-house FDTD MATLAB code takes into account time variations of the permittivity.
{"title":"Wave Propagation in Metallic Slab Waveguides Undergoing Arbitrary Temporal Variations of Permittivity","authors":"Mahdi Chegnizadeh, M. Memarian, K. Mehrany","doi":"10.1109/IranianCEE.2019.8786763","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786763","url":null,"abstract":"A general and closed form solution of wave propagation in a metallic slab waveguide undergoing any arbitrary, time-varying function of permittivity is presented. The formulation is general and accurate to account for all temporal variations of permittivity. Parametric amplification and frequency conversion is demonstrated using this closed form formulation. Moreover, all results are validated and confirmed using full wave FDTD simulation. The in-house FDTD MATLAB code takes into account time variations of the permittivity.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"16 1","pages":"1529-1533"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82590022","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786509
E. Jafari, T. Binazadeh
Solving the robust regulation problem is considered for singular systems in the presence of actuator saturation, model uncertainties, and disturbances. Although singularity of system equations, caused by existence of algebraic equations together with differential ones, gives it a great ability for describing more precise behaviors of dynamical systems, it also causes complexities in analysis and design procedures. On the other hand, considering actuator saturation constraint as a practical limitation for the system needs precise mathematical calculations. In this regard, a theorem is given that shows the proposed robust control law is capable to drive the output of the system to track a step reference in presence of uncertainties, disturbances, and actuator saturation constraint. Investigating different cases of saturation function with precise mathematical operations in the given theorem based on Lyapunov approach analysis demonstrates that the regulation goal is achieved even if the input saturation takes place. The accuracy of the proposed control law is finally verified by computer simulations.
{"title":"Robust Nonlinear Control of Singular Systems with Saturating Actuator Constraint","authors":"E. Jafari, T. Binazadeh","doi":"10.1109/IranianCEE.2019.8786509","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786509","url":null,"abstract":"Solving the robust regulation problem is considered for singular systems in the presence of actuator saturation, model uncertainties, and disturbances. Although singularity of system equations, caused by existence of algebraic equations together with differential ones, gives it a great ability for describing more precise behaviors of dynamical systems, it also causes complexities in analysis and design procedures. On the other hand, considering actuator saturation constraint as a practical limitation for the system needs precise mathematical calculations. In this regard, a theorem is given that shows the proposed robust control law is capable to drive the output of the system to track a step reference in presence of uncertainties, disturbances, and actuator saturation constraint. Investigating different cases of saturation function with precise mathematical operations in the given theorem based on Lyapunov approach analysis demonstrates that the regulation goal is achieved even if the input saturation takes place. The accuracy of the proposed control law is finally verified by computer simulations.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"28 1","pages":"1113-1117"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82256645","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786705
Reyhane Gazori, G. Mirjalily
Routing is an important issue in vehicular ad-hoc networks due to the high speed movement of vehicles in a hostile dynamic environment. In order to improve the performance of existing routing protocols, in this paper a geographic routing protocol based on the backbone structure is introduced. In the proposed algorithm, the traffic lights are used as bridges for routing assessment instead of using the mobile vehicles. This prevents the network failures in the route selection process. In proposed protocol, by evaluating the paths through exchanging control packets between bridge nodes, the path with the smallest hop count and maximum connectivity is selected. The simulation results show that by using the proposed method, significant improvement in the network performance is obtained in comparison to the similar protocol named SCRP.
{"title":"SBGRP as an Improved Stable CDS-Based Routing Protocol in Vehicular Ad Hoc Networks","authors":"Reyhane Gazori, G. Mirjalily","doi":"10.1109/IranianCEE.2019.8786705","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786705","url":null,"abstract":"Routing is an important issue in vehicular ad-hoc networks due to the high speed movement of vehicles in a hostile dynamic environment. In order to improve the performance of existing routing protocols, in this paper a geographic routing protocol based on the backbone structure is introduced. In the proposed algorithm, the traffic lights are used as bridges for routing assessment instead of using the mobile vehicles. This prevents the network failures in the route selection process. In proposed protocol, by evaluating the paths through exchanging control packets between bridge nodes, the path with the smallest hop count and maximum connectivity is selected. The simulation results show that by using the proposed method, significant improvement in the network performance is obtained in comparison to the similar protocol named SCRP.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"21 1","pages":"1979-1983"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76523546","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786367
S. Hemmati, Mostafa Barigh
In transport applications such as high-speed trains and hybrid machines, it is essential for motor to achieve speeds above nominal speed and operate in the field weakening region. A new approach to weaken the permanent magnet machine field is proposed in this paper. In this way, the three-phase stator winding of motor is divided into two parts, each part being connected to an inverter. This method is called the double winding switching technique, which can surely multiply the field weakening speed range of permanent magnetized motors. This research examines the stator windings connections, allowing the motor to work at normal and high speeds at constant output power.
{"title":"A new approach for field weakening in a surface mounted permanent magnet synchronous motor by winding switching *","authors":"S. Hemmati, Mostafa Barigh","doi":"10.1109/IranianCEE.2019.8786367","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786367","url":null,"abstract":"In transport applications such as high-speed trains and hybrid machines, it is essential for motor to achieve speeds above nominal speed and operate in the field weakening region. A new approach to weaken the permanent magnet machine field is proposed in this paper. In this way, the three-phase stator winding of motor is divided into two parts, each part being connected to an inverter. This method is called the double winding switching technique, which can surely multiply the field weakening speed range of permanent magnetized motors. This research examines the stator windings connections, allowing the motor to work at normal and high speeds at constant output power.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"344 1","pages":"509-514"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76570958","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786422
A. Ahmadi, Hossein Ghasemian, E. Abiri, R. Ghasemi
In this paper, a novel 10-bit 1.2 GS/s hybrid digital to analog converter (DAC) is presented. In this work, a resistor ladder beside the unit current sources is employed to implement the higher value bits. Using a resistor ladder for scaling the current sources causes to decrease the number of current sources impressively. In addition, the complicated decoder circuits used in segmented structures are replaced by simple digital logics with fewer transistors. simulation results in 65 nm CMOS technology demonstrate that the proposed DAC consumes only 8.4 mW power with a 1.2 V power supply, which is an improvement in comparison with recent works. Moreover, simulation results show that the INL and DNL are less than 0.25 and 0.12 LSB, respectively. Also, more than 74 dB SFDR is obtained in the entire 600 MHz Nyquist bandwidth.
{"title":"A Novel 10-bit 1.2GS/s Hybrid Digital to Analog Converter with 8.4mW Power Consumption","authors":"A. Ahmadi, Hossein Ghasemian, E. Abiri, R. Ghasemi","doi":"10.1109/IranianCEE.2019.8786422","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786422","url":null,"abstract":"In this paper, a novel 10-bit 1.2 GS/s hybrid digital to analog converter (DAC) is presented. In this work, a resistor ladder beside the unit current sources is employed to implement the higher value bits. Using a resistor ladder for scaling the current sources causes to decrease the number of current sources impressively. In addition, the complicated decoder circuits used in segmented structures are replaced by simple digital logics with fewer transistors. simulation results in 65 nm CMOS technology demonstrate that the proposed DAC consumes only 8.4 mW power with a 1.2 V power supply, which is an improvement in comparison with recent works. Moreover, simulation results show that the INL and DNL are less than 0.25 and 0.12 LSB, respectively. Also, more than 74 dB SFDR is obtained in the entire 600 MHz Nyquist bandwidth.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"16 1","pages":"163-168"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87642667","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786388
Nader Asadi, M. Eftekhari
Deep neural networks are state-of-art models in computer vision and image recognition problems. However, it is shown that these models are highly vulnerable to intentionally perturbed inputs named adversarial examples. This problem has attracted a lot of attention in recent years. In this paper, a novel approach is proposed for generating adversarial examples by perturbing latent representation of an input image that causes to mislead trained classifier network. Also, it is shown that perturbing dense representation of image results in transforming key features of it with respect to classification task. Our experimental results show that this slight transformation in the features of the image can easily fool the classifier network. We also show the impact of adding perturbations with the large magnitude to the corresponding generated adversarial example.
{"title":"A Novel Image Perturbation Approach: Perturbing Latent Representation","authors":"Nader Asadi, M. Eftekhari","doi":"10.1109/IranianCEE.2019.8786388","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786388","url":null,"abstract":"Deep neural networks are state-of-art models in computer vision and image recognition problems. However, it is shown that these models are highly vulnerable to intentionally perturbed inputs named adversarial examples. This problem has attracted a lot of attention in recent years. In this paper, a novel approach is proposed for generating adversarial examples by perturbing latent representation of an input image that causes to mislead trained classifier network. Also, it is shown that perturbing dense representation of image results in transforming key features of it with respect to classification task. Our experimental results show that this slight transformation in the features of the image can easily fool the classifier network. We also show the impact of adding perturbations with the large magnitude to the corresponding generated adversarial example.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"47 1","pages":"1895-1899"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87816895","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 : 2019-04-01DOI: 10.1109/IranianCEE.2019.8786401
P. Ahmadvand, R. Ebrahimpour, Payam Ahmadvand
Face identification has been one of the most challenging and attractive areas, and it has become a popular research area in the computer vision community because of its impact in application areas, such as surveillance, security systems, and biometrics. With the introduction deep learning technique, face recognition accuracy has skyrocketed. The state-of-the-art has achieved close to 100% recognition rates by applying convolutional neural networks (CNN), as these networks are more robust in regards to image variation than other methods are. However, in an uncontrolled environment with illumination, extreme poses, and variations in facial expression, the face recognition problem is far from being solved. This paper has two contributions, the first one is to discuss real-world challenges in face identification and suggest solutions for them. To overcome one these challenges, we train an MLP model to validate the output of CNN which reduce the false positive rate. The second contribution is to implement an OpenVX computational graph model to achieve real-time performance and reduce the run-time.
{"title":"Deep Real-world and Real-time Face Identification System","authors":"P. Ahmadvand, R. Ebrahimpour, Payam Ahmadvand","doi":"10.1109/IranianCEE.2019.8786401","DOIUrl":"https://doi.org/10.1109/IranianCEE.2019.8786401","url":null,"abstract":"Face identification has been one of the most challenging and attractive areas, and it has become a popular research area in the computer vision community because of its impact in application areas, such as surveillance, security systems, and biometrics. With the introduction deep learning technique, face recognition accuracy has skyrocketed. The state-of-the-art has achieved close to 100% recognition rates by applying convolutional neural networks (CNN), as these networks are more robust in regards to image variation than other methods are. However, in an uncontrolled environment with illumination, extreme poses, and variations in facial expression, the face recognition problem is far from being solved. This paper has two contributions, the first one is to discuss real-world challenges in face identification and suggest solutions for them. To overcome one these challenges, we train an MLP model to validate the output of CNN which reduce the false positive rate. The second contribution is to implement an OpenVX computational graph model to achieve real-time performance and reduce the run-time.","PeriodicalId":6683,"journal":{"name":"2019 27th Iranian Conference on Electrical Engineering (ICEE)","volume":"38 1","pages":"1989-1993"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86838531","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：