摘要 :
By the experimental strengthening method in working conditions, the war characteristics resulting from applying slurry film lubrication or omitting them were investigated. The results show that the wear rate reduces significantly ...
展开
By the experimental strengthening method in working conditions, the war characteristics resulting from applying slurry film lubrication or omitting them were investigated. The results show that the wear rate reduces significantly when slurry film lubrication in applied. The wear resulting from the new technique is 53%-83% lower than that for the original boat at the sliding speed of 3.5-5.0km/h and the pressure of 2600-5700Pa. It has been confirmed that the sliding condition were improved greatly because of the slurry-film lubrication.
收起
摘要 :
By the experimental strengthening method in working conditions, the war characteristics resulting from applying slurry film lubrication or omitting them were investigated. The results show that the wear rate reduces significantly ...
展开
By the experimental strengthening method in working conditions, the war characteristics resulting from applying slurry film lubrication or omitting them were investigated. The results show that the wear rate reduces significantly when slurry film lubrication in applied. The wear resulting from the new technique is 53%-83% lower than that for the original boat at the sliding speed of 3.5-5.0km/h and the pressure of 2600-5700Pa. It has been confirmed that the sliding condition were improved greatly because of the slurry-film lubrication.
收起
摘要 :
When learning to sketch, beginners start with simple and flexible shapes, and then gradually strive for more complex and accurate ones in the subsequent training sessions. In this paper, we design a "shape curriculum" for learning...
展开
When learning to sketch, beginners start with simple and flexible shapes, and then gradually strive for more complex and accurate ones in the subsequent training sessions. In this paper, we design a "shape curriculum" for learning continuous Signed Distance Function (SDF) on shapes, namely Curriculum DeepSDF. Inspired by how humans learn, Curriculum DeepSDF organizes the learning task in ascending order of difficulty according to the following two criteria: surface accuracy and sample difficulty. The former considers stringency in supervising with ground truth, while the latter regards the weights of hard training samples near complex geometry and fine structure. More specifically, Curriculum DeepSDF learns to reconstruct coarse shapes at first, and then gradually increases the accuracy and focuses more on complex local details. Experimental results show that a carefully-designed curriculum leads to significantly better shape reconstructions with the same training data, training epochs and network architecture as DeepSDF. We believe that the application of shape curricula can benefit the training process of a wide variety of 3D shape representation learning methods.
收起
摘要 :
When learning to sketch, beginners start with simple and flexible shapes, and then gradually strive for more complex and accurate ones in the subsequent training sessions. In this paper, we design a "shape curriculum" for learning...
展开
When learning to sketch, beginners start with simple and flexible shapes, and then gradually strive for more complex and accurate ones in the subsequent training sessions. In this paper, we design a "shape curriculum" for learning continuous Signed Distance Function (SDF) on shapes, namely Curriculum DeepSDF. Inspired by how humans learn, Curriculum DeepSDF organizes the learning task in ascending order of difficulty according to the following two criteria: surface accuracy and sample difficulty. The former considers stringency in supervising with ground truth, while the latter regards the weights of hard training samples near complex geometry and fine structure. More specifically, Curriculum DeepSDF learns to reconstruct coarse shapes at first, and then gradually increases the accuracy and focuses more on complex local details. Experimental results show that a carefully-designed curriculum leads to significantly better shape reconstructions with the same training data, training epochs and network architecture as DeepSDF. We believe that the application of shape curricula can benefit the training process of a wide variety of 3D shape representation learning methods.
收起
摘要 :
Nowdays, it is prevalent to train deep learning (DL) models in cloud-native platforms that actively leverage containerization and orchestration technologies for high elasticity, low and flexible operation cost, and many other bene...
展开
Nowdays, it is prevalent to train deep learning (DL) models in cloud-native platforms that actively leverage containerization and orchestration technologies for high elasticity, low and flexible operation cost, and many other benefits. However, it also faces new challenges and our work is focusing on those related to I/O throughput for training, including complex data access with complicated performance tuning, lack of cache capacity with specialized hardware to match its high and dynamic I/O requirement, and inefficient I/O resource sharing across different training jobs. We propose Fluid, a cloud-native platform that provides DL training jobs with a data abstraction called Fluid Dataset to access training data from heterogeneous sources in a unified manner with transparent and elastic data acceleration powered by auto-tuned cache runtimes. In addition, it comes with an on-the-fly cache system autoscaler that can intelligently scale up and down the cache capacity to match the online training speed of each individual DL job. To improve the overall performance of multiple DL jobs, Fluid can co-orchestrate the data cache and DL jobs by arranging job scheduling in an appropriate order. Our experimental results show significant performance improvement of each individual DL job which uses dynamic computing resources with Fluid. In addition, for scheduling multiple DL jobs with same datasets, Fluid gives around 2x performance speedup when integrated with existing widely-used and cutting-edge scheduling solutions. Fluid is now an open source project hosted by Cloud Native Computing Foundation (CNCF) with adopters in production including Alibaba Cloud, Tencent Cloud, Weibo.com, China Telecom, etc.
收起
摘要 :
Partially-premixed low-temperature combustion avoids the soot and NO_x generation area on the Ф-T diagram to reduce both engine NO_x and soot emissions. Compared with the HCCI combustion mode, partially-premixed combustion (PPC) ...
展开
Partially-premixed low-temperature combustion avoids the soot and NO_x generation area on the Ф-T diagram to reduce both engine NO_x and soot emissions. Compared with the HCCI combustion mode, partially-premixed combustion (PPC) has better combustion controllability. The purpose of controlling the combustion phase can be achieved by adjusting injection timing and strategy. Based on a 4 cylinder turbocharged diesel engine, this paper aims at investigating the influence of injection strategy to the engine combustion and emission formation under the condition of single injection and split injection PPC strategy respectively, in which the primary purpose focus on the emission characteristics of particles. Results show that the early-injection PPC formed by single injection can reduce the quantity, quality and geometric mean diameter (GMD) of particles obviously. However, when the injection timing is too early, the number of particles will rise to the level as the traditional mode since the concentration of particulate matter in the nucleation region increases dramatically. The pilot injection strategy can reduce soot emission significantly, while at the same time, the concentration of nucleation mode particles remains stable, which differs from single early injection PPC. However, if the timing of the pilot injection advanced to a certain value, the quantity, quality and GMD of particles are almost unchanged. Increasing the proportion of pilot injection fuel to form a higher proportion of premixed combustion, both the quantity and quality of particles will decrease, GMD will undergo an increase-decrease process. CO and HC emissions are very low under the early injection PPC. While in the pre-injection strategy, the fuel burning rate is low and many intermediate products generate, so the CO and HC emission will increase as the pre-injection fuel ratio increases and pre-injection timing advances.
收起
摘要 :
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in...
展开
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in its application to range proof. We demonstrate that their batch proof and verification technique causes a concern such that in some cases a malicious prover without the claimed knowledge may pass the verification. As a result their range proof scheme to prove that a secret committed integer is in an interval range is not so reliable and cannot guarantee that the committed integer is in the range in some special cases. To ease the concern, we employ an efficient membership proof technique to replace the batch proof and verification technique in their range proof scheme and re-design it to achieve the claimed high efficiency with practical small ranges.
收起
摘要 :
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in...
展开
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in its application to range proof. We demonstrate that their batch proof and verification technique causes a concern such that in some cases a malicious prover without the claimed knowledge may pass the verification. As a result their range proof scheme to prove that a secret committed integer is in an interval range is not so reliable and cannot guarantee that the committed integer is in the range in some special cases. To ease the concern, we employ an efficient membership proof technique to replace the batch proof and verification technique in their range proof scheme and re-design it to achieve the claimed high efficiency with practical small ranges.
收起
摘要 :
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in...
展开
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in its application to range proof. We demonstrate that their batch proof and verification technique causes a concern such that in some cases a malicious prover without the claimed knowledge may pass the verification. As a result their range proof scheme to prove that a secret committed integer is in an interval range is not so reliable and cannot guarantee that the committed integer is in the range in some special cases. To ease the concern, we employ an efficient membership proof technique to replace the batch proof and verification technique in their range proof scheme and re-design it to achieve the claimed high efficiency with practical small ranges.
收起
摘要 :
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in...
展开
A batch proof and verification technique is employed to design efficient range proof with practical small ranges in AFRICACRYPT 2010. It is shown in this paper that the batch proof and verification technique is not always sound in its application to range proof. We demonstrate that their batch proof and verification technique causes a concern such that in some cases a malicious prover without the claimed knowledge may pass the verification. As a result their range proof scheme to prove that a secret committed integer is in an interval range is not so reliable and cannot guarantee that the committed integer is in the range in some special cases. To ease the concern, we employ an efficient membership proof technique to replace the batch proof and verification technique in their range proof scheme and re-design it to achieve the claimed high efficiency with practical small ranges.
收起