Browsing by Author "Lin, Wen-Hui"

Now showing 1 - 5 of 5

A Cross-age Face Recognition Approach Using Fog Computing Architecture for User Authentication on Mobile Devices
(2018-01-01)
Lin, Wen-Hui
;
Wang, Ping
;
Wu, Bao-Hua
;
Chao, Kuo-Ming
;
Lo, Chi-Chun
Mobile commerce security needs quality services where mobile devices respond in real-time and traditional cloud computing approach uses a centralized architecture does not support these systems with such time dependency well. Thus, this study proposes a cross-age face-recognition model with MobileNets to determine the identity of a user in the situation of aging appearance change on a fog computing architecture. Especially, this study uses a face-recognition technique based on a distributed concept. In this concept, MobileNet-based face recognition system effectively performs cross-age face recognition for identity authentication with the enhanced facial features of the users locally. Overall, the MobileNet model reaches acceptable prediction accuracy with lower latency in face recognition process compared with that of the existing CNNs schemes.
A Face-Recognition Approach Using Deep Reinforcement Learning Approach for User Authentication
(2017-01-01)
Wang, Ping
;
Lin, Wen-Hui
;
Chao, Kuo-Ming
;
Lo, Chi-Chun
;
資訊管理與財務金融系註：原資管所+財金所
Numerous crime-related security concerns exist in e-commerce transactions recently. User authentication for mobile payment has numerous approaches including face recognition, iris scan, and fingerprint scan to identify user's true identity by comparing the biometric features of users with patterns in the signature database. Existing studies on the face recognition problem focus mainly on the static analysis to determine the face recognition precision by examining the facial features of images with different facial expressions for users rather than the dynamic aspects where images were are often vague affected by lighting changes with different poses. Because the lighting, facial expressions, and facial details varied in the face recognition process. Consequently, it limits the effectiveness of scheme with which to determine the true identity. Accordingly, this study focused on a face recognition process under the situation of vague facial features using deep reinforcement learning (DRL) approach with convolutional neuron networks (CNNs) thru facial feature extraction, transformation, and comparison to determine the user identity for mobile payment. Specifically, the proposed authentication scheme uses back propagation algorithm to effectively improve the accuracy of face recognition using feed-forward network architecture for CNNs. Overall, the proposed scheme provided a higher precision of face recognition (100% at gamma correction gamma located in [0.5, 1.6]) compared with the average precision for face image (approximately 99.5% at normal lighting gamma = 1) of the existing CNN schemes with ImageNet 2012 Challenge training data set.
An Efficient Flow Control Approach for SDN-based Network Threat Detection and Migration Using Support Vector Machine
(2016)
Wang, Ping
;
Lin, Hsiao-Chung
;
Lin, Wen-Hui
;
Chao, Kuo-Ming
;
Lo, Chi-Chun
Most existing approaches for solving the network threat problems focus on the specific security mechanisms, for example, network intrusion detection system (NIDS) detection, firewall configuration, rather than on flow management approaches to defend network threats with an SDN (Software Defined Networking) architecture. Accordingly, this study proposes an improved behaviour-based SVM (support vector machine) with learning algorithm for use in the security monitoring system (SMS) to categorize network threats for network intrusion detection system. The model also adopted the ID3 decision tree theory to outrank raw features and determine the most qualified features to train support vector classifier (SVC) considering the overall detection precision rate of experiments which speeds up the learning of normal and intrusive patterns and and increases the accuracy of detecting intrusion. By using sFlow collector and analyzer associated with sFlow-RT toolset, the experimental results proved that the SMS enables a defender to classify the network threats with defence strategies and defend network threats.
Using Dynamic Taint Approach for of Malware Threat
(2015)
Wang, Ping
;
Lin, Wen-Hui
;
Chao, Wun Jie
;
Chao, Kuo-Ming
;
Lo, Chi-Chun
Most existing approaches focus on examining the values are dangerous for information flow within inter-suspicious modules of cloud applications (apps) in a host by using malware threat analysis, rather than the risk posed by suspicious apps were connected to the cloud computing server. Accordingly, this paper proposes a taint propagation analysis model incorporating a weighted spanning tree analysis scheme to track data with taint marking using several taint checking tools. In the proposed model, Android programs perform dynamic taint propagation to analyse the spread of and risks posed by suspicious apps were connected to the cloud computing server. In determining the risk of taint propagation, risk and defence capability are used for each taint path for assisting a defender in recognising the attack results against network threats caused by malware infection and estimate the losses of associated taint sources. Finally, a case of threat analysis of a typical cyber security attack is presented to demonstrate the proposed approach. Our approach verified the details of an attack sequence for malware infection by incorporating a finite state machine (FSM) to appropriately reflect the real situations at various configuration settings and safeguard deployment. The experimental results proved that the threat analysis model allows a defender to convert the spread of taint propagation to loss and practically estimate the risk of a specific threat by using behavioural analysis with real malware infection.
Using malware for software-defined networking-based smart home security management through a taint checking approach
(2016-08-01)
Wang, Ping
;
Chao, Kuo-Ming
;
Lo, Chi-Chun
;
Lin, Wen-Hui
;
Lin, Hsiao-Chung
Numerous security concerns exist in smart home systems in which Internet of Things devices are connected through a home network to enable control using a centralised gateway with a handset device from the Internet. Safeguarding personal information privacy is an increasing concern in smart living services. To guarantee the mobile security of smart living services, security managers use taint checking approaches with dynamic taint propagation analysis operations to examine how a software-defined networking app uses sensitive information and investigate suspicious security vulnerabilities of devices and the effects of the spread of taint propagation over the Internet by identifying taint paths. For solving the dynamic taint propagation analysis problem, most approaches focus on cloud computing applications (apps) with malware threat analysis that involves program vulnerability analyses, rather than on the risk posed by suspicious apps connected to the cloud computing server. Accordingly, this article proposes a taint propagation analysis model incorporating a weighted spanning tree analysis scheme for tracking data with taint marking using several taint checking tools with an open software-defined networking architecture for solving the dynamic taint propagation analysis problem. In the proposed model, Android programs perform dynamic taint propagation to analyse the spread of risks posed by suspicious apps connected to the centralised gateway in a smart home system. In probabilistic risk analysis, risk and defence capability are used for each taint path to assist a defender in recognising the attack results against network threats caused by malware infection and to estimate the losses of associated taint sources. A case of threat analysis of a typical cyber security attack is presented to demonstrate the proposed approach. A new approach was used for verifying the details of an attack sequence for malware infection by incorporating a finite state machine to appropriately represent the real dynamic taint propagation analysis situations at various configuration settings and safeguard deployment. The experimental results proved that the threat analysis model enables a defender to convert the spread of taint propagation to loss and estimate the risk of a specific threat using behavioural analysis associated with 60 families of real malware. Consequently, our scheme was significantly effective in predicting the risk and loss of tainted data propagation for security concerns in smart home systems when the number of taint paths associated with the propagation rules discovered through taint analysis was increased.