We introduce the use of geodesic distances and geodesic radius for calculating pairwise similarities needed in various graph cuts based methods. By using geodesics on an edge strength function we are able to calculate...
详细信息
We introduce the use of geodesic distances and geodesic radius for calculating pairwise similarities needed in various graph cuts based methods. By using geodesics on an edge strength function we are able to calculate similarities between pixels in a more natural way. Our technique improves the speed and reliability of calculating similarities and leads to reasonably good image segmentation results. Our algorithm takes an edge strength function as its input and its speed is independent of the feature dimension or the distance measure used.
Graph analytics has emerged as one of the fundamental techniques to support modern Internet applications. As real-world graph data is generated and stored globally, the scale of the graph that needs to be processed ke...
详细信息
ISBN:
(纸本)9781538641293
Graph analytics has emerged as one of the fundamental techniques to support modern Internet applications. As real-world graph data is generated and stored globally, the scale of the graph that needs to be processed keeps growing. It is critical to efficiently process graphs across multiple geographically distributed datacenters, running wide-area graph analytics. Existing graph analytics frameworks are not designed to run across multiple datacenters well, as they implement a Bulk Synchronous Parallel model that requires excessive wide-area data transfers. In this paper, we present a new Hierarchical Synchronous Parallel model designed and implemented for synchronization across datacenters with a much improved efficiency in inter-datacenter communication. Our new model requires no modifications to graph analytics applications, yet guarantees their convergence and correctness. Our prototype implementation on Apache Spark can achieve up to 32% lower WAN bandwidth usage, 49% faster convergence, and 30% less total cost for benchmark graph algorithms, with input data stored across five geographically distributed datacenters.
Despite remarkable progress on face recognition, little attention has been given to robustly recognize expression variant faces from a single sample image per person. One way to deal with the recognition of faces unde...
详细信息
Despite remarkable progress on face recognition, little attention has been given to robustly recognize expression variant faces from a single sample image per person. One way to deal with the recognition of faces under above conditions is by using local statistical approaches which appear to be more robust against variations in facial expression. In this paper, we propose a new weighted matching method based on our recent work of AWPPZMA to recognize expression variant faces when only one exemplar image per enrolled subject is available. The proposed weighting method gives more significance to those parts of the face with facial expression variations that change less compared to neutral face image and less significance to those parts that change more. In this contribution, we use the difference between local area in the input face and its corresponding local area in the neutral face image as a measure of observable structure changes. The encouraging experimental results demonstrate that the proposed method provides a new solution to the problem of robustly recognizing expression variant faces in single model databases.
For geographic routing in 2D ad hoc networks, greedy algorithm is efficient if no holes exist. However, greedy forwarding fails when a message reaches a local-minimum. Thus, face routing is used to solve the problem o...
详细信息
For geographic routing in 2D ad hoc networks, greedy algorithm is efficient if no holes exist. However, greedy forwarding fails when a message reaches a local-minimum. Thus, face routing is used to solve the problem of local-minimums. Unfortunately, few research results can be applied to geographic routing in 3D environments. This paper proposes an algorithm GSG (Greedy Surface routing Greedy), for geographic routing in 3D environments. GSG firstly partitions whole network with 3D Restricted Delau-nay Triangulation. When a local-minimum is encountered because of a hole, a message is transferred on a surface with triangles in order to jump out of a local-minimum area. Triangles and isolated edges are defined as 3D components on Surface. By means of identifying intersecting triangles and edges, efficient routes are constructed on surfaces by bypassing local-minimums. For sparsely deployed networks, an algorithm DFS based on 3D RNG is provided. Simulation results show that GSG outperforms the algorithms such as GRG and GHG.
NAND flash memory is widely used for secondary storage today. The flash translation layer (FTL) is the embedded software that is responsible for managing and operating in flash storage system. One important module of ...
详细信息
ISBN:
(纸本)9781450321532
NAND flash memory is widely used for secondary storage today. The flash translation layer (FTL) is the embedded software that is responsible for managing and operating in flash storage system. One important module of the FTL performs RAM management. It is well-known to have a significant impact on flash storage system's performance. This paper proposes an efficient RAM management scheme called TreeFTL. As the name suggests, TreeFTL organizes address translation pages and data pages in RAM in a tree structure, through which it dynamically adapts to workloads by adjusting the partitions for address mapping and data buffering. TreeFTL also employs a lightweight mechanism to implement the least recently used (LRU) algorithm for RAM cache evictions. Experiments show that compared to the two latest schemes for RAM management in flash storage system, TreeFTL can reduce service time by 46.6% and 49.0% on average, respectively, with a 64MB RAM cache.
Schema matching plays an important role in many database applications, such as ontology merging, data integration, data warehouse and dataspaces. The problem of schema matching is to find the semantic correspondence b...
详细信息
ISBN:
(纸本)9781467352512
Schema matching plays an important role in many database applications, such as ontology merging, data integration, data warehouse and dataspaces. The problem of schema matching is to find the semantic correspondence between attributes of schemas to be matched. In this paper, we propose multi-schema matching based on clustering techniques. Traditional matching techniques mainly address matching tasks between two attributes, namely pairwise-attribute correspondence. However, there exist lots of applications that require the semantic correspondence among multiple attributes. Thus, we will focus on matching multiple attributes, which is more difficult than pairwise-attribute correspondence. We employ the clustering techniques to solve the multi-schema matching problem. We use the well-known TFIDF weighting method to convert each attribute in schemas to a point in the vector space model. Then, these attributes can be partitioned into different clusters each of which has a specific semantics topic. Finally, the attributes partitioned into the same cluster are similar with higher confidence. We validate our approach with an experimental study, the results of which demonstrate that our approach is effective and has good performance.
Connectivity is a fundamental requirement for Wireless Sensor Networks (WSNs). Failures in multiple nodes can disrupt network connectivity, leading to the formation of disjointed subnetworks. To restore network connec...
详细信息
ISBN:
(数字)9798350387780
ISBN:
(纸本)9798350387797
Connectivity is a fundamental requirement for Wireless Sensor Networks (WSNs). Failures in multiple nodes can disrupt network connectivity, leading to the formation of disjointed subnetworks. To restore network connectivity, a UAV-assisted Adaptive Relay nodes Placement algorithm (UARP). is proposed in this paper. To concentrate relay nodes placement and reduce the travel distance of UAVs, a concept of UAV Areas is introduced to identify the areas between interrupted network partitions. Based on the number of network partitions within UAV Areas, two different strategies are employed to place relay nodes and reconnect the partitions for the scenarios with more and fewer partitions, respectively. With the proposed algorithm, relay nodes are strategically placed within UAV Areas, which can reduce the required number of relay nodes and shorten the UAV travel distance. Simulation results indicate that the UARP algorithm reduces relay usage by 10% to 20% and decreases UAV travel distance by 5% to 10% compared to other algorithms. The algorithm also demonstrates satisfactory performance in terms of packet delivery rate, convergence time, and transmission delay.
In this paper, we present a novel algorithm that automatically counts rebar on production line based on video analysis. Our algorithm allows that conveyor belt moves at rather higher speed and rebars roll forward or b...
详细信息
In this paper, we present a novel algorithm that automatically counts rebar on production line based on video analysis. Our algorithm allows that conveyor belt moves at rather higher speed and rebars roll forward or backward on conveyor belt. There are mainly two parts in our proposed algorithm. Firstly, a method is developed to divide the video frame into three partitions, and then a set of heuristic rules are deduced based on these partitions. Secondly, for cases excluded by above rules, a sequence matching algorithm is developed to identify repeated sequence of rebar between adjacent frames, and then number of newly arrived rebar in every frame can be computed. The experimental results show that our method can support high-speed and precise rebar counting.
In this paper, we present a novel texture classification algorithm inspired by the self-assembling behavior of real ants when building live structures with their bodies. The proposed algorithm employs dyadic Gabor fil...
详细信息
In this paper, we present a novel texture classification algorithm inspired by the self-assembling behavior of real ants when building live structures with their bodies. The proposed algorithm employs dyadic Gabor filter banks for extracting discriminant features from images containing multiple textures not known to the algorithm. The feature space is clustered using the novel ant tree clustering (ATC) algorithm based on the similarity of ants carrying the feature vectors. The results thus obtained show promise of the proposed approach.
Understanding the community structure in graphs arising from complex network is an important and difficult problem, both from theoretical and practical points of views. Although a lot of community detection algorithms...
详细信息
Understanding the community structure in graphs arising from complex network is an important and difficult problem, both from theoretical and practical points of views. Although a lot of community detection algorithms have been proposed in the last decade, there is still no satisfactory way to determine if a given network possesses or not a community structure, that is, its nodes can be partitioned in well separated clusters. In this paper, we propose a new criterion based on the study of the formation of consensual communities obtained by running several times a non deterministic algorithm. By testing on synthetic benchmarks (with known structure) and on several real world networks, we show that the graphs can be categorized in several classes according to the dynamic of the consensual communities formation process. This result is promising to derive new approaches to characterize the modular structure of graphs.
暂无评论