摘要 :
The rapid development of point cloud learning has driven point cloud completion into a new era. However, the information flows of most existing completion methods are solely feedforward, and high-level information is rarely reused...
展开
The rapid development of point cloud learning has driven point cloud completion into a new era. However, the information flows of most existing completion methods are solely feedforward, and high-level information is rarely reused to improve low-level feature learning. To this end, we propose a novel Feedback Network (FBNet) for point cloud completion, in which present features are efficiently refined by rerouting subsequent fine-grained ones. Firstly, partial inputs are fed to a Hierarchical Graph-based Network (HGNet) to generate coarse shapes. Then, we cascade several Feedback-Aware Completion (FBAC) Blocks and unfold them across time recurrently. Feedback connections between two adjacent time steps exploit fine-grained features to improve present shape generations. The main challenge of building feedback connections is the dimension mismatching between present and subsequent features. To address this, the elaborately designed point Cross Transformer exploits efficient information from feedback features via cross attention strategy and then refines present features with the enhanced feedback features. Quantitative and qualitative experiments on several datasets demonstrate the superiority of proposed FBNet compared to state-of-the-art methods on point completion task.
收起
摘要 :
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bu...
展开
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bursting", the requests may be rejected because of the limited resources of private cloud. In this paper, we propose the online cost-aware service requests scheduling strategy in hybrid clouds (OCS) which could make suitable requests placement decisions real-time and minimize the cost of renting public cloud resources with a low rate of rejected requests. All service requests are divided into two categories, the special requests ask to be accepted on private cloud, and the normal requests are insensitive on private or public cloud. In addition, all requests arrive in random, without any prior knowledge of future arrivals. We transform the online model into a one-shot optimization problem by taking advantage of Lyapunov optimization techniques, then employ the optimal decay algorithm to solve the one-shot problem. The simulation results demonstrate that OCS is trade-off between cost and rejection rate, meanwhile it can let the resource utilization arbitrarily close to the optimum.
收起
摘要 :
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bu...
展开
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bursting", the requests may be rejected because of the limited resources of private cloud. In this paper, we propose the online cost-aware service requests scheduling strategy in hybrid clouds (OCS) which could make suitable requests placement decisions real-time and minimize the cost of renting public cloud resources with a low rate of rejected requests. All service requests are divided into two categories, the special requests ask to be accepted on private cloud, and the normal requests are insensitive on private or public cloud. In addition, all requests arrive in random, without any prior knowledge of future arrivals. We transform the online model into a one-shot optimization problem by taking advantage of Lyapunov optimization techniques, then employ the optimal decay algorithm to solve the one-shot problem. The simulation results demonstrate that OCS is trade-off between cost and rejection rate, meanwhile it can let the resource utilization arbitrarily close to the optimum.
收起
摘要 :
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bu...
展开
The hybrid cloud computing model has been attracting considerable attention in the past years. Due to security and controllability of private cloud, some special requests ask to be scheduled on private cloud, when requests are "bursting", the requests may be rejected because of the limited resources of private cloud. In this paper, we propose the online cost-aware service requests scheduling strategy in hybrid clouds (OCS) which could make suitable requests placement decisions real-time and minimize the cost of renting public cloud resources with a low rate of rejected requests. All service requests are divided into two categories, the special requests ask to be accepted on private cloud, and the normal requests are insensitive on private or public cloud. In addition, all requests arrive in random, without any prior knowledge of future arrivals. We transform the online model into a one-shot optimization problem by taking advantage of Lyapunov optimization techniques, then employ the optimal decay algorithm to solve the one-shot problem. The simulation results demonstrate that OCS is trade-off between cost and rejection rate, meanwhile it can let the resource utilization arbitrarily close to the optimum.
收起
摘要 :
The academia domestic and abroad has been caring that whether moral education can improve students' moral decision,and in vestigating the methods of moral education.In China,much research has explored the current situation of prof...
展开
The academia domestic and abroad has been caring that whether moral education can improve students' moral decision,and in vestigating the methods of moral education.In China,much research has explored the current situation of professional moral education in the area of accounting major.It is shown that the accounting moral education is ig nored,compared with the basic knowledge and skills.This paper will make a comparative experiment of professional moral education in the au diting course of accounting major,to find the influences of moral educa tion on the abilities to recognize moral issues,the first stage in the moral decision model.It is found that,to a high extent,moral education im proves the students' abilities to recognize moral issues.Additionally,the methods of case study and discussions are more effective than the intro duction of professional moral standards only.
收起
摘要 :
The academia domestic and abroad has been caring that whether moral education can improve students' moral decision,and in vestigating the methods of moral education.In China,much research has explored the current situation of...
展开
The academia domestic and abroad has been caring that whether moral education can improve students' moral decision,and in vestigating the methods of moral education.In China,much research has explored the current situation of professional moral education in the area of accounting major.It is shown that the accounting moral education is ig nored,compared with the basic knowledge and skills.This paper will make a comparative experiment of professional moral education in the au diting course of accounting major,to find the influences of moral educa tion on the abilities to recognize moral issues,the first stage in the moral decision model.It is found that,to a high extent,moral education im proves the students' abilities to recognize moral issues.Additionally,the methods of case study and discussions are more effective than the intro duction of professional moral standards only.
收起
摘要 :
Multiversion software transactional memory (STM) allows a transaction to read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This abi...
展开
Multiversion software transactional memory (STM) allows a transaction to read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This ability to "commit in the past" is particularly appealing for long-running read-only transactions, which may otherwise starve in many STM systems, because short-running peers modify data out from under them before they have a chance to finish. Most previous approaches to multiversioning have been designed as an integral part of some larger STM system, and have assumed an object-oriented, garbage-collected language. We describe, instead, how multiversioning may be implemented on top of an almost arbitrary "word-based" STM system. To the best of our knowledge, ours is the first work (for any kind of STM) to combine bounded space consumption with guaranteed wait freedom for read-only transactions (in the form presented here, it may require writers to be blocking). We make no assumptions about data or metadata layout, though we do require that the base system provide a hash function with certain ordering properties. We neither require nor interfere with automatic garbage collection. Privatization safety can be ensured - without compromising wait freedom for readers - either by forcing privatizing writers to wait for all extant readers or by requiring that programmers explicitly identify the data being privatized.
收起
摘要 :
Multiversion software transactional memory (STM) allows a transaction to, read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This ab...
展开
Multiversion software transactional memory (STM) allows a transaction to, read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This ability to "commit in the past" is particularly appealing for long-running read-only transactions, which may otherwise starve in many STM systems, because short-running peers modify data out from under them before they have a chance to finish. Most previous approaches to multiversioning have been designed as an integral part of some larger STM system, and have assumed an object-oriented, garbage-collected language. We describe, instead, how multiversioning may be implemented on top of an almost arbitrary "word-based" STM system. To the best of our knowledge, ours is the first work (for any kind of STM) to combine bounded space consumption with guaranteed wait freedom for read-only transactions (in the form presented here, it may require writers to be blocking). We make no assumptions about data or metadata layout, though we do require that the base system provide a hash function with certain ordering properties. We neither require nor interfere with automatic garbage collection. Privatization safety can be ensured-without compromising wait freedom for readers-either by forcing privatizing writers to wait for all extant readers or by requiring that programmers explicitly identify the data being privatized.
收起
摘要 :
Multiversion software transactional memory (STM) allows a transaction to read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This abi...
展开
Multiversion software transactional memory (STM) allows a transaction to read old values of a recently updated object, after which the transaction may serialize before transactions that committed earlier in physical time. This ability to "commit in the past" is particularly appealing for long-running read-only transactions, which may otherwise starve in many STM systems, because short-running peers modify data out from under them before they have a chance to finish. Most previous approaches to multiversioning have been designed as an integral part of some larger STM system, and have assumed an object-oriented, garbage-collected language. We describe, instead, how multiversioningmay be implemented on top of an almost arbitrary "wordbased" STM system. To the best of our knowledge, ours is the first work (for any kind of STM) to combine bounded space consumptionwith guaranteed wait freedom for read-only transactions (in the form presented here, it may require writers to be blocking). We make no assumptions about data or metadata layout, though we do require that the base system provide a hash function with certain ordering properties. We neither require nor interfere with automatic garbage collection. Privatization safety can be ensured--without compromisingwait freedom for readers--either by forcing privatizing writers to wait for all extant readers or by requiring that programmers explicitly identify the data being privatized.
收起
摘要 :
Deterministic parallelism has become an increasingly attractive concept: a deterministic parallel program may be easier to construct, debug, understand, and maintain. However, there exist many different definitions of "determinism...
展开
Deterministic parallelism has become an increasingly attractive concept: a deterministic parallel program may be easier to construct, debug, understand, and maintain. However, there exist many different definitions of "determinism" for parallel programming. Many existing definitions have not yet been fully formalized, and the relationships among these definitions are still unclear. We argue that formalism is needed, and that history-based semantics—as used, for example, to define the Java and C++ memory models—provides a useful lens through which to view the notion of determinism. As a first step, we suggest several history-based definitions of determinism. We discuss some of their comparative advantages, prove containment relationships among them, and identify programming idioms that ensure them. We also propose directions for future work.
收起
原文获取