Latent factor models based on matrix/tensor factorization techniques are widely recognized in recommender ***,matrix/tensor factorization techniques are computationally intensive,which greatly limits their usages in r...
详细信息
ISBN:
(纸本)9781467349994
Latent factor models based on matrix/tensor factorization techniques are widely recognized in recommender ***,matrix/tensor factorization techniques are computationally intensive,which greatly limits their usages in real world *** address this problem,we improve the classic matrix factorization models by establishing fuzzy mapping relationships between contexts and latent factors,and making use of contextual information to initialize user/item feature *** results on standard test set MovieLens 1M show that the proposed algorithms could achieve far better prediction accuracy while reducing the iteration number by 25%.
VANET is new kind of Ad hoc. It is widely used in ITS (Intelligent Transportation systems), which has many characteristics as large scale network, fast moving nodes, the frequently changing topological structure and e...
详细信息
ISBN:
(纸本)9781467362498;9781467362481
VANET is new kind of Ad hoc. It is widely used in ITS (Intelligent Transportation systems), which has many characteristics as large scale network, fast moving nodes, the frequently changing topological structure and easily divided networks. Therefore, routing protocol design must fully consider of this characteristics with much node information that bring a great challenge to the security of VANET. In this paper, we divide the information type into four categories based on different content, and analyze the security threats in different information type, then summarize the existing security technologies and give the possible research directions.
In order to enhance the extensibility of current attack feature extracted anddetection means for App-ddoS(Application Layerdistributeddenial of Service, App-ddoS)attacks, a novel feature extracted method based on m...
详细信息
A video copy detection method based on fusion of spatio-temporal features is proposed in this paper. Firstly, trajectories are built and lens boundaries are detected by SUrF features analyzing, then normalized histogr...
详细信息
In 3d image processing, such as medical volume data and industrial CT volume data analysis, speckle noise suppressing significantly affects their accuracy. This paper utilizes a group of volume morphology arithmetic o...
详细信息
The bottleneck of NetFlow is the memory *** solve this problem,we propose continuous reading and writing method whose efficiency reaches 90% or *** single parallel multi-linked lists method used to handle the inactive...
详细信息
ISBN:
(纸本)9781467349994
The bottleneck of NetFlow is the memory *** solve this problem,we propose continuous reading and writing method whose efficiency reaches 90% or *** single parallel multi-linked lists method used to handle the inactive timeout flows is based on the continuous reading and writing *** analysis and experimental results show that the single parallel multi-linked lists method can be applied in OC-768(40Gbps) to manage tens of millions of entries.
Image feature extraction is an important technology in image matching andretrieval. For the problem of high computational complexity of spatial domain image feature extraction using the SIFT algorithm, and by studyin...
详细信息
In 3d image processing,such as medical volume data and industrial CT volume data analysis,speckle noise suppressing significantly affects their *** paper utilizes a group of volume morphology arithmetic operators,main...
详细信息
In 3d image processing,such as medical volume data and industrial CT volume data analysis,speckle noise suppressing significantly affects their *** paper utilizes a group of volume morphology arithmetic operators,mainly including volume open and volume close,by extending area morphology into 3-d *** these operators,the light anddark objects of small size could be removeddirectly from the 3-d space of the target objects,while the connectivity of the main 3-d target objects in the volume data is still *** improve the volume morphology operations efficiency,we decompose volume data in bitplanes instead of in gray scale space to reduce the binary *** verify the effect of the improved volume morphology operators,they are applied to suppress speckle noises in 3-d images of coral andrat skull and compared with original volume morphology *** results show that the algorithm proposed in this paper can significantly reduce computing time,while maintaining comparable results to original operations results.
In medical image processing,interactive image segmentation is an important part,because it can obtain accurate segment results with less human effort compared with manual *** proposed an improved algorithm of maximal ...
详细信息
In medical image processing,interactive image segmentation is an important part,because it can obtain accurate segment results with less human effort compared with manual *** proposed an improved algorithm of maximal similarity basedregion merging,compared with the algorithm proposed in [3],our algorithm use SLIC superpixels segmentation to obtain presegmentedregions,using SLIC superpixles,it is easy to control the number of presegmentation *** also introduce the texture features differeces while rigion merging,so we can obtain the accuracy of similarity *** results show that our algorithm can obtain comparable results.
Many images suffer from poor *** is necessary to enhance the contrast of *** enhancement improves an image appearance by increasing dominance of some features or by decreasing ambiguity between different regions of th...
详细信息
Many images suffer from poor *** is necessary to enhance the contrast of *** enhancement improves an image appearance by increasing dominance of some features or by decreasing ambiguity between different regions of the *** equalization(HE) is widely used for contrast enhancement in a variety of applications due to its simple function and ***,it usually results in excessive contrast enhancement,which causes the unnatural look and visual artifacts of the processed *** improved algorithms,such as Brightness preserving Bi-Histogram Equalization,recursive Mean-Separate Histogram Equalization,dual Sub-Image Histogram Equalization and so on,are devoted to preserving the input image brightness and will be disappointed when the input image has low luminance.A novel method based on histogram similarity is developed to overcome the drawbacks of the classic HE algorithms for gray scale *** to some of the conventional HE methods,the proposed method produces better contrast and image quality.
暂无评论