Recent years, deep convolutional neural network has led to significant improvements in face recognition and becomes one of the most popular techniques in computer vision community. However, deep CNN model requires vas...
详细信息
ISBN:
(数字)9781728160245
ISBN:
(纸本)9781728160252
Recent years, deep convolutional neural network has led to significant improvements in face recognition and becomes one of the most popular techniques in computer vision community. However, deep CNN model requires vast amounts of data and time for training and deploying. To solve this problem, we present a deep compact convolutional neural network for face representation. First we apply PCA for initializing convolution filter. Then we adopt DCT and binary hashing for extracting face features. The models are trained on the CASIA-webFace datasets under Caffe framework. Experimental results show that the proposed methods achieve competitive accuracy on the LFW verification benchmark.
Deep learning (DL) is widely used in radio frequency fingerprint identification (RFFI). However, in few-shot case, traditional DL-based RFFI need to construct auxiliary dataset to realize radio frequency fingerprint i...
详细信息
Session-based recommendation tries to make use of anonymous session data to deliver high-quality recommendation under the condition that user-profiles and the complete historical behavioral data of a target user are u...
详细信息
The difficult of detecting, response, tracing the malicious behavior in cloud has brought great challenges to the law enforcement in combating cybercrimes. This paper presents a malicious behavior oriented framework o...
详细信息
With the development of science and technology in recent years,the planning and construction of smart cities have entered a new *** them,smart city safety management is the foundation that supports the stable developm...
详细信息
With the development of science and technology in recent years,the planning and construction of smart cities have entered a new *** them,smart city safety management is the foundation that supports the stable development of the entire city,and location-based services are one of its important technical *** popularity of GPS-equipped devices provides a large amount of data for trajectory *** personal semantic locations by mining these data is an applicationbased on location *** semantic locations are frequently visited by individual users and have significant semantic meaning to users（such as home,work place,etc.）.The discovery of the user’s personal semantic location involves obtaining the physical location and *** present,related research mostly uses clustering algorithms to obtain the physical location,of which the DBSCAN algorithm is the most commonly *** the traditional DBSCAN algorithm determines whether a sample belongs to a certain cluster,it will generate a large number of repeated calculations,which will reduce the program operation *** order to solve this problem,this paper proposes an improved DBSCAN algorithm *** algorithm uses a k-d tree to screen samples that need to be *** excluding samples that do not need to be calculated,the purpose of improving the calculation efficiency can be *** the problem of semantic location recognition,this paper pre-defines the daily behavior patterns of people,and identifies the location semantics based on the temporal characteristics of the samples in each *** calculating the distance between samples,the algorithm proposed here is 8.8 times more efficient than the traditional algorithm.
The difficult of detecting, response, tracing the malicious behavior in cloud has brought great challenges to the law enforcement in combating cybercrimes. This paper presents a malicious behavior oriented framework o...
详细信息
The difficult of detecting, response, tracing the malicious behavior in cloud has brought great challenges to the law enforcement in combating cybercrimes. This paper presents a malicious behavior oriented framework of detection, emergency response, traceability, and digital forensics in cloud environment. A cloud-based malicious behavior detection mechanism based on SDN is constructed, which implements full-traffic flow detection technology and malicious virtual machine detection based on memory analysis. The emergency response and traceability module can clarify the types of the malicious behavior and the impacts of the events, and locate the source of the event. The key nodes and paths of the infection topology or propagation path of the malicious behavior will be located security measure will be dispatched timely. The proposed IaaS service based forensics module realized the virtualization facility memory evidence extraction and analysis techniques, which can solve volatile data loss problems that often happened in traditional forensic methods.
暂无评论