Pub Date : 2023-12-01DOI: 10.1007/s11390-022-1887-1
Abstract
Synthesizing garment dynamics according to body motions is a vital technique in computer graphics. Physics-based simulation depends on an accurate model of the law of kinetics of cloth, which is time-consuming, hard to implement, and complex to control. Existing data-driven approaches either lack temporal consistency, or fail to handle garments that are different from body topology. In this paper, we present a motion-inspired real-time garment synthesis workflow that enables high-level control of garment shape. Given a sequence of body motions, our workflow is able to generate corresponding garment dynamics with both spatial and temporal coherence. To that end, we develop a transformerbased garment synthesis network to learn the mapping from body motions to garment dynamics. Frame-level attention is employed to capture the dependency of garments and body motions. Moreover, a post-processing procedure is further taken to perform penetration removal and auto-texturing. Then, textured clothing animation that is collision-free and temporally-consistent is generated. We quantitatively and qualitatively evaluated our proposed workflow from different aspects. Extensive experiments demonstrate that our network is able to deliver clothing dynamics which retain the wrinkles from the physics-based simulation, while running 1 000 times faster. Besides, our workflow achieved superior synthesis performance compared with alternative approaches. To stimulate further research in this direction, our code will be publicly available soon.
{"title":"Motion-Inspired Real-Time Garment Synthesis with Temporal-Consistency","authors":"","doi":"10.1007/s11390-022-1887-1","DOIUrl":"https://doi.org/10.1007/s11390-022-1887-1","url":null,"abstract":"<h3>Abstract</h3> <p>Synthesizing garment dynamics according to body motions is a vital technique in computer graphics. Physics-based simulation depends on an accurate model of the law of kinetics of cloth, which is time-consuming, hard to implement, and complex to control. Existing data-driven approaches either lack temporal consistency, or fail to handle garments that are different from body topology. In this paper, we present a motion-inspired real-time garment synthesis workflow that enables high-level control of garment shape. Given a sequence of body motions, our workflow is able to generate corresponding garment dynamics with both spatial and temporal coherence. To that end, we develop a transformerbased garment synthesis network to learn the mapping from body motions to garment dynamics. Frame-level attention is employed to capture the dependency of garments and body motions. Moreover, a post-processing procedure is further taken to perform penetration removal and auto-texturing. Then, textured clothing animation that is collision-free and temporally-consistent is generated. We quantitatively and qualitatively evaluated our proposed workflow from different aspects. Extensive experiments demonstrate that our network is able to deliver clothing dynamics which retain the wrinkles from the physics-based simulation, while running 1 000 times faster. Besides, our workflow achieved superior synthesis performance compared with alternative approaches. To stimulate further research in this direction, our code will be publicly available soon.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"9 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139656281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-01DOI: 10.1007/s11390-022-1919-x
Abstract
Agile hardware design is gaining increasing momentum and bringing new chips in larger quantities to the market faster. However, it also takes new challenges for compiler developers to retarget existing compilers to these new chips in shorter time than ever before. Currently, retargeting a compiler backend, e.g., an LLVM backend to a new target, requires compiler developers to write manually a set of target description files (totalling 10 300+ lines of code (LOC) for RISC-V in LLVM), which is error-prone and time-consuming. In this paper, we introduce a new approach, Automatic Target Description File Generation (ATG), which accelerates the generation of a compiler backend for a new target by generating its target description files automatically. Given a new target, ATG proceeds in two stages. First, ATG synthesizes a small list of target-specific properties and a list of code-layout templates from the target description files of a set of existing targets with similar instruction set architectures (ISAs). Second, ATG requests compiler developers to fill in the information for each instruction in the new target in tabular form according to the list of target-specific properties synthesized and then generates its target description files automatically according to the list of code-layout templates synthesized. The first stage can often be reused by different new targets sharing similar ISAs. We evaluate ATG using nine RISC-V instruction sets drawn from a total of 1 029 instructions in LLVM 12.0. ATG enables compiler developers to generate compiler backends for these ISAs that emit the same assembly code as the existing compiler backends for RISC-V but with significantly less development effort (by specifying each instruction in terms of up to 61 target-specific properties only).
{"title":"Automatic Target Description File Generation","authors":"","doi":"10.1007/s11390-022-1919-x","DOIUrl":"https://doi.org/10.1007/s11390-022-1919-x","url":null,"abstract":"<h3>Abstract</h3> <p>Agile hardware design is gaining increasing momentum and bringing new chips in larger quantities to the market faster. However, it also takes new challenges for compiler developers to retarget existing compilers to these new chips in shorter time than ever before. Currently, retargeting a compiler backend, e.g., an LLVM backend to a new target, requires compiler developers to write manually a set of target description files (totalling 10 300+ lines of code (LOC) for RISC-V in LLVM), which is error-prone and time-consuming. In this paper, we introduce a new approach, Automatic Target Description File Generation (ATG), which accelerates the generation of a compiler backend for a new target by generating its target description files automatically. Given a new target, ATG proceeds in two stages. First, ATG synthesizes a small list of target-specific properties and a list of code-layout templates from the target description files of a set of existing targets with similar instruction set architectures (ISAs). Second, ATG requests compiler developers to fill in the information for each instruction in the new target in tabular form according to the list of target-specific properties synthesized and then generates its target description files automatically according to the list of code-layout templates synthesized. The first stage can often be reused by different new targets sharing similar ISAs. We evaluate ATG using nine RISC-V instruction sets drawn from a total of 1 029 instructions in LLVM 12.0. ATG enables compiler developers to generate compiler backends for these ISAs that emit the same assembly code as the existing compiler backends for RISC-V but with significantly less development effort (by specifying each instruction in terms of up to 61 target-specific properties only).</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"155-156 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139656170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-01DOI: 10.1007/s11390-023-1346-7
Abstract
Local differential privacy (LDP) approaches to collecting sensitive information for frequent itemset mining (FIM) can reliably guarantee privacy. Most current approaches to FIM under LDP add “padding and sampling” steps to obtain frequent itemsets and their frequencies because each user transaction represents a set of items. The current state-of-the-art approach, namely set-value itemset mining (SVSM), must balance variance and bias to achieve accurate results. Thus, an unbiased FIM approach with lower variance is highly promising. To narrow this gap, we propose an Item-Level LDP frequency oracle approach, named the Integrated-with-Hadamard-Transform-Based Frequency Oracle (IHFO). For the first time, Hadamard encoding is introduced to a set of values to encode all items into a fixed vector, and perturbation can be subsequently applied to the vector. An FIM approach, called optimized united itemset mining (O-UISM), is proposed to combine the padding-and-sampling-based frequency oracle (PSFO) and the IHFO into a framework for acquiring accurate frequent itemsets with their frequencies. Finally, we theoretically and experimentally demonstrate that O-UISM significantly outperforms the extant approaches in finding frequent itemsets and estimating their frequencies under the same privacy guarantee.
摘要 为频繁项集挖掘(FIM)收集敏感信息的局部差分隐私(LDP)方法可以可靠地保证隐私。由于每笔用户交易都代表一组项目,因此目前大多数 LDP 下的频繁项集挖掘方法都增加了 "填充和采样 "步骤,以获得频繁项集及其频率。目前最先进的方法,即集值项集挖掘(SVSM),必须在方差和偏差之间取得平衡,才能获得准确的结果。因此,一种无偏见、方差较小的 FIM 方法大有可为。为了缩小这一差距,我们提出了一种项级 LDP 频率甲骨文方法,名为基于哈达玛德变换的集成频率甲骨文(IHFO)。我们首次在一组值中引入哈达玛编码,将所有项目编码为一个固定的向量,随后可对该向量进行扰动。我们提出了一种称为优化联合项集挖掘(O-UISM)的 FIM 方法,将基于填充和采样的频率神谕(PSFO)和 IHFO 结合到一个框架中,以获取精确的频繁项集及其频率。最后,我们通过理论和实验证明,O-UISM 在寻找频繁项集和估算其频率方面明显优于现有方法。
{"title":"Hadamard Encoding Based Frequent Itemset Mining under Local Differential Privacy","authors":"","doi":"10.1007/s11390-023-1346-7","DOIUrl":"https://doi.org/10.1007/s11390-023-1346-7","url":null,"abstract":"<h3>Abstract</h3> <p>Local differential privacy (LDP) approaches to collecting sensitive information for frequent itemset mining (FIM) can reliably guarantee privacy. Most current approaches to FIM under LDP add “padding and sampling” steps to obtain frequent itemsets and their frequencies because each user transaction represents a set of items. The current state-of-the-art approach, namely set-value itemset mining (SVSM), must balance variance and bias to achieve accurate results. Thus, an unbiased FIM approach with lower variance is highly promising. To narrow this gap, we propose an Item-Level LDP frequency oracle approach, named the Integrated-with-Hadamard-Transform-Based Frequency Oracle (IHFO). For the first time, Hadamard encoding is introduced to a set of values to encode all items into a fixed vector, and perturbation can be subsequently applied to the vector. An FIM approach, called optimized united itemset mining (O-UISM), is proposed to combine the padding-and-sampling-based frequency oracle (PSFO) and the IHFO into a framework for acquiring accurate frequent itemsets with their frequencies. Finally, we theoretically and experimentally demonstrate that O-UISM significantly outperforms the extant approaches in finding frequent itemsets and estimating their frequencies under the same privacy guarantee.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"13 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139659401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-30DOI: 10.1007/s11390-023-1420-1
Wen-Yu Gao, Hang Gao
Cluster deletion and strong triadic closure are two important NP-complete problems that have received significant attention due to their applications in various areas, including social networks and data analysis. Although cluster deletion and strong triadic closure are closely linked by induced paths on three vertices, there are subtle differences between them. In some cases, the solutions of strong triadic closure and cluster deletion are quite different. In this paper, we study the parameterized algorithms for these two problems. More specifically, we focus on the kernels of these two problems. Instead of separating the critical clique and its neighbors for analysis, we consider them as a whole, which allows us to more effectively bound the number of related vertices. In addition, in analyzing the kernel of strong triadic closure, we introduce the concept of edge-disjoint induced path on three vertices, which enables us to obtain the lower bound of weak edge number in a more concise way. Our analysis demonstrates that cluster deletion and strong triadic closure both admit 2k-vertex kernels. These results represent improvements over previously best-known kernels for both problems. Furthermore, our analysis provides additional insights into the relationship between cluster deletion and strong triadic closure.
{"title":"2k-Vertex Kernels for Cluster Deletion and Strong Triadic Closure","authors":"Wen-Yu Gao, Hang Gao","doi":"10.1007/s11390-023-1420-1","DOIUrl":"https://doi.org/10.1007/s11390-023-1420-1","url":null,"abstract":"<p>Cluster deletion and strong triadic closure are two important NP-complete problems that have received significant attention due to their applications in various areas, including social networks and data analysis. Although cluster deletion and strong triadic closure are closely linked by induced paths on three vertices, there are subtle differences between them. In some cases, the solutions of strong triadic closure and cluster deletion are quite different. In this paper, we study the parameterized algorithms for these two problems. More specifically, we focus on the kernels of these two problems. Instead of separating the critical clique and its neighbors for analysis, we consider them as a whole, which allows us to more effectively bound the number of related vertices. In addition, in analyzing the kernel of strong triadic closure, we introduce the concept of edge-disjoint induced path on three vertices, which enables us to obtain the lower bound of weak edge number in a more concise way. Our analysis demonstrates that cluster deletion and strong triadic closure both admit 2<i>k</i>-vertex kernels. These results represent improvements over previously best-known kernels for both problems. Furthermore, our analysis provides additional insights into the relationship between cluster deletion and strong triadic closure.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"24 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139656204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-30DOI: 10.1007/s11390-023-1295-1
Ming He, Yan Chen, Hong-Ke Zhao, Qi Liu, Le Wu, Yu Cui, Gui-Hua Zeng, Gui-Quan Liu
Automatic generation of Chinese classical poetry is still a challenging problem in artificial intelligence. Recently, Encoder-Decoder models have provided a few viable methods for poetry generation. However, by reviewing the prior methods, two major issues still need to be settled: 1) most of them are one-stage generation methods without further polishing; 2) they rarely take into consideration the restrictions of poetry, such as tone and rhyme. Intuitively, some ancient Chinese poets tended first to write a coarse poem underlying aesthetics and then deliberated its semantics; while others first create a semantic poem and then refine its aesthetics. On this basis, in order to better imitate the human creation procedure of poems, we propose a two-stage method (i.e., restricted polishing generation method) of which each stage focuses on the different aspects of poems (i.e., semantics and aesthetics), which can produce a higher quality of generated poems. In this way, the two-stage method develops into two symmetrical generation methods, the aesthetics-to-semantics method and the semantics-to-aesthetics method. In particular, we design a sampling method and a gate to formulate the tone and rhyme restrictions, which can further improve the rhythm of the generated poems. Experimental results demonstrate the superiority of our proposed two-stage method in both automatic evaluation metrics and human evaluation metrics compared with baselines, especially in yielding consistent improvements in tone and rhyme.
{"title":"Composing Like an Ancient Chinese Poet: Learn to Generate Rhythmic Chinese Poetry","authors":"Ming He, Yan Chen, Hong-Ke Zhao, Qi Liu, Le Wu, Yu Cui, Gui-Hua Zeng, Gui-Quan Liu","doi":"10.1007/s11390-023-1295-1","DOIUrl":"https://doi.org/10.1007/s11390-023-1295-1","url":null,"abstract":"<p>Automatic generation of Chinese classical poetry is still a challenging problem in artificial intelligence. Recently, Encoder-Decoder models have provided a few viable methods for poetry generation. However, by reviewing the prior methods, two major issues still need to be settled: 1) most of them are one-stage generation methods without further polishing; 2) they rarely take into consideration the restrictions of poetry, such as tone and rhyme. Intuitively, some ancient Chinese poets tended first to write a coarse poem underlying aesthetics and then deliberated its semantics; while others first create a semantic poem and then refine its aesthetics. On this basis, in order to better imitate the human creation procedure of poems, we propose a two-stage method (i.e., restricted polishing generation method) of which each stage focuses on the different aspects of poems (i.e., semantics and aesthetics), which can produce a higher quality of generated poems. In this way, the two-stage method develops into two symmetrical generation methods, the aesthetics-to-semantics method and the semantics-to-aesthetics method. In particular, we design a sampling method and a gate to formulate the tone and rhyme restrictions, which can further improve the rhythm of the generated poems. Experimental results demonstrate the superiority of our proposed two-stage method in both automatic evaluation metrics and human evaluation metrics compared with baselines, especially in yielding consistent improvements in tone and rhyme.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"118 1 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657219","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cognitive diagnosis is an important issue of intelligent education systems, which aims to estimate students’ proficiency on specific knowledge concepts. Most existing studies rely on the assumption of static student states and ignore the dynamics of proficiency in the learning process, which makes them unsuitable for online learning scenarios. In this paper, we propose a unified temporal item response theory (UTIRT) framework, incorporating temporality and randomness of proficiency evolving to get both accurate and interpretable diagnosis results. Specifically, we hypothesize that students’ proficiency varies as a Wiener process and describe a probabilistic graphical model in UTIRT to consider temporality and randomness factors. Furthermore, based on the relationship between student states and exercising answers, we hypothesize that the answering result at time k contributes most to inferring a student's proficiency at time k, which also reflects the temporality aspect and enables us to get analytical maximization (M-step) in the expectation maximization (EM) algorithm when estimating model parameters. Our UTIRT is a framework containing unified training and inferencing methods, and is general to cover several typical traditional models such as Item Response Theory (IRT), multidimensional IRT (MIRT), and temporal IRT (TIRT). Extensive experimental results on real-world datasets show the effectiveness of UTIRT and prove its superiority in leveraging temporality theoretically and practically over TIRT.
认知诊断是智能教育系统的一个重要问题,其目的是估计学生对特定知识概念的掌握程度。现有研究大多依赖于学生静态状态的假设,忽略了学习过程中能力的动态变化,因此不适合在线学习场景。在本文中,我们提出了一个统一的时序项目反应理论(UTIRT)框架,将能力演进的时序性和随机性结合起来,以获得既准确又可解释的诊断结果。具体来说,我们假设学生的能力变化是一个维纳过程,并在UTIRT中描述了一个概率图形模型,以考虑时间性和随机性因素。此外,基于学生状态和练习答案之间的关系,我们假设第 k 次的答题结果对推断学生在第 k 次的能力贡献最大,这也反映了时间性的一面,并使我们能够在估计模型参数时在期望最大化(EM)算法中获得解析最大化(M-step)。我们的UTIRT是一个包含统一训练和推断方法的框架,它具有通用性,可涵盖多个典型的传统模型,如项目反应理论(IRT)、多维IRT(MIRT)和时间IRT(TIRT)。在实际数据集上的大量实验结果表明了UTIRT的有效性,并证明了它在理论和实践上都比TIRT更善于利用时间性。
{"title":"A Probabilistic Framework for Temporal Cognitive Diagnosis in Online Learning Systems","authors":"Jia-Yu Liu, Fei Wang, Hai-Ping Ma, Zhen-Ya Huang, Qi Liu, En-Hong Chen, Yu Su","doi":"10.1007/s11390-022-1332-5","DOIUrl":"https://doi.org/10.1007/s11390-022-1332-5","url":null,"abstract":"<p>Cognitive diagnosis is an important issue of intelligent education systems, which aims to estimate students’ proficiency on specific knowledge concepts. Most existing studies rely on the assumption of static student states and ignore the dynamics of proficiency in the learning process, which makes them unsuitable for online learning scenarios. In this paper, we propose a unified temporal item response theory (UTIRT) framework, incorporating temporality and randomness of proficiency evolving to get both accurate and interpretable diagnosis results. Specifically, we hypothesize that students’ proficiency varies as a Wiener process and describe a probabilistic graphical model in UTIRT to consider temporality and randomness factors. Furthermore, based on the relationship between student states and exercising answers, we hypothesize that the answering result at time <i>k</i> contributes most to inferring a student's proficiency at time <i>k</i>, which also reflects the temporality aspect and enables us to get analytical maximization (M-step) in the expectation maximization (EM) algorithm when estimating model parameters. Our UTIRT is a framework containing unified training and inferencing methods, and is general to cover several typical traditional models such as Item Response Theory (IRT), multidimensional IRT (MIRT), and temporal IRT (TIRT). Extensive experimental results on real-world datasets show the effectiveness of UTIRT and prove its superiority in leveraging temporality theoretically and practically over TIRT.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"39 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-30DOI: 10.1007/s11390-021-1251-x
Wan-Rong Gao, Jian-Bin Fang, Chun Huang, Chuan-Fu Xu, Zheng Wang
Cache performance is a critical design constraint for modern many-core systems. Since the cache often works in a “black-box” manner, it is difficult for the software to reason about the cache behavior to match the running software to the underlying hardware. To better support code optimization, we need to understand and characterize the cache behavior. While cache performance characterization is heavily studied on traditional x86 architectures, there is little work for understanding the cache implementations on emerging ARMv8-based many-cores. This paper presents a comprehensive study to evaluate the cache architecture design on three representative ARMv8 multi-cores, Phytium 2000+, ThunderX2, and Kunpeng 920 (KP920). To this end, we develop wrBench, a micro-benchmark suite to measure the realized latency and bandwidth of caches at different memory hierarchies when performing core-to-core communication. Our evaluation provides inter-core latency and bandwidth in different cache levels and coherency states for the three ARMv8 many-cores. The quantitative performance data is shown in tables. We mine the characteristics of caches and coherency protocols by analyzing the data for the three processors, Phytium 2000+, ThunderX2, and KP920. Our paper also provides discussions and guidelines for optimizing memory access on ARMv8 many-cores.
{"title":"wrBench: Comparing Cache Architectures and Coherency Protocols on ARMv8 Many-Core Systems","authors":"Wan-Rong Gao, Jian-Bin Fang, Chun Huang, Chuan-Fu Xu, Zheng Wang","doi":"10.1007/s11390-021-1251-x","DOIUrl":"https://doi.org/10.1007/s11390-021-1251-x","url":null,"abstract":"<p>Cache performance is a critical design constraint for modern many-core systems. Since the cache often works in a “black-box” manner, it is difficult for the software to reason about the cache behavior to match the running software to the underlying hardware. To better support code optimization, we need to understand and characterize the cache behavior. While cache performance characterization is heavily studied on traditional x86 architectures, there is little work for understanding the cache implementations on emerging ARMv8-based many-cores. This paper presents a comprehensive study to evaluate the cache architecture design on three representative ARMv8 multi-cores, Phytium 2000+, ThunderX2, and Kunpeng 920 (KP920). To this end, we develop wrBench, a micro-benchmark suite to measure the realized latency and bandwidth of caches at different memory hierarchies when performing core-to-core communication. Our evaluation provides inter-core latency and bandwidth in different cache levels and coherency states for the three ARMv8 many-cores. The quantitative performance data is shown in tables. We mine the characteristics of caches and coherency protocols by analyzing the data for the three processors, Phytium 2000+, ThunderX2, and KP920. Our paper also provides discussions and guidelines for optimizing memory access on ARMv8 many-cores.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"1 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657221","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-30DOI: 10.1007/s11390-022-1320-9
Han-Bo Zhang, Peng Wang, Ming-Ming Zhang, Wei Wang
In the last decade, there has been significant progress in time series classification. However, in real-world industrial settings, it is expensive and difficult to obtain high-quality labeled data. Therefore, the positive and unlabeled learning (PU-learning) problem has become more and more popular recently. The current PU-learning approaches of the time series data suffer from low accuracy due to the lack of negative labeled time series. In this paper, we propose a novel shapelet based two-step (2STEP) PU-learning approach. In the first step, we generate shapelet features based on the positive time series, which are used to select a set of negative examples. In the second step, based on both positive and negative time series, we select the final features and build the classification model. The experimental results show that our 2STEP approach can improve the average F1 score on 15 datasets by 9.1% compared with the baselines, and achieves the highest F1 score on 10 out of 15 time series datasets.
{"title":"Shapelet Based Two-Step Time Series Positive and Unlabeled Learning","authors":"Han-Bo Zhang, Peng Wang, Ming-Ming Zhang, Wei Wang","doi":"10.1007/s11390-022-1320-9","DOIUrl":"https://doi.org/10.1007/s11390-022-1320-9","url":null,"abstract":"<p>In the last decade, there has been significant progress in time series classification. However, in real-world industrial settings, it is expensive and difficult to obtain high-quality labeled data. Therefore, the positive and unlabeled learning (PU-learning) problem has become more and more popular recently. The current PU-learning approaches of the time series data suffer from low accuracy due to the lack of negative labeled time series. In this paper, we propose a novel shapelet based two-step (2STEP) PU-learning approach. In the first step, we generate shapelet features based on the positive time series, which are used to select a set of negative examples. In the second step, based on both positive and negative time series, we select the final features and build the classification model. The experimental results show that our 2STEP approach can improve the average <i>F</i>1 score on 15 datasets by 9.1% compared with the baselines, and achieves the highest <i>F</i>1 score on 10 out of 15 time series datasets.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"285 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-30DOI: 10.1007/s11390-022-1479-0
Gui-Rong Bai, Qing-Bin Liu, Shi-Zhu He, Kang Liu, Jun Zhao
Although neural approaches have yielded state-of-the-art results in the sentence matching task, their performance inevitably drops dramatically when applied to unseen domains. To tackle this cross-domain challenge, we address unsupervised domain adaptation on sentence matching, in which the goal is to have good performance on a target domain with only unlabeled target domain data as well as labeled source domain data. Specifically, we propose to perform self-supervised tasks to achieve it. Different from previous unsupervised domain adaptation methods, self-supervision can not only flexibly suit the characteristics of sentence matching with a special design, but also be much easier to optimize. When training, each self-supervised task is performed on both domains simultaneously in an easy-to-hard curriculum, which gradually brings the two domains closer together along the direction relevant to the task. As a result, the classifier trained on the source domain is able to generalize to the unlabeled target domain. In total, we present three types of self-supervised tasks and the results demonstrate their superiority. In addition, we further study the performance of different usages of self-supervised tasks, which would inspire how to effectively utilize self-supervision for cross-domain scenarios.
{"title":"Unsupervised Domain Adaptation on Sentence Matching Through Self-Supervision","authors":"Gui-Rong Bai, Qing-Bin Liu, Shi-Zhu He, Kang Liu, Jun Zhao","doi":"10.1007/s11390-022-1479-0","DOIUrl":"https://doi.org/10.1007/s11390-022-1479-0","url":null,"abstract":"<p>Although neural approaches have yielded state-of-the-art results in the sentence matching task, their performance inevitably drops dramatically when applied to unseen domains. To tackle this cross-domain challenge, we address unsupervised domain adaptation on sentence matching, in which the goal is to have good performance on a target domain with only unlabeled target domain data as well as labeled source domain data. Specifically, we propose to perform self-supervised tasks to achieve it. Different from previous unsupervised domain adaptation methods, self-supervision can not only flexibly suit the characteristics of sentence matching with a special design, but also be much easier to optimize. When training, each self-supervised task is performed on both domains simultaneously in an easy-to-hard curriculum, which gradually brings the two domains closer together along the direction relevant to the task. As a result, the classifier trained on the source domain is able to generalize to the unlabeled target domain. In total, we present three types of self-supervised tasks and the results demonstrate their superiority. In addition, we further study the performance of different usages of self-supervised tasks, which would inspire how to effectively utilize self-supervision for cross-domain scenarios.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"15 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657374","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The scarcity of bilingual parallel corpus imposes limitations on exploiting the state-of-the-art supervised translation technology. One of the research directions is employing relations among multi-modal data to enhance performance. However, the reliance on manually annotated multi-modal datasets results in a high cost of data labeling. In this paper, the topic semantics of images is proposed to alleviate the above problem. First, topic-related images can be automatically collected from the Internet by search engines. Second, topic semantics is sufficient to encode the relations between multi-modal data such as texts and images. Specifically, we propose a visual topic semantic enhanced translation (VTSE) model that utilizes topic-related images to construct a cross-lingual and cross-modal semantic space, allowing the VTSE model to simultaneously integrate the syntactic structure and semantic features. In the above process, topic similar texts and images are wrapped into groups so that the model can extract more robust topic semantics from a set of similar images and then further optimize the feature integration. The results show that our model outperforms competitive baselines by a large margin on the Multi30k and the Ambiguous COCO datasets. Our model can use external images to bring gains to translation, improving data efficiency.
{"title":"Visual Topic Semantic Enhanced Machine Translation for Multi-Modal Data Efficiency","authors":"Chao Wang, Si-Jia Cai, Bei-Xiang Shi, Zhi-Hong Chong","doi":"10.1007/s11390-023-1302-6","DOIUrl":"https://doi.org/10.1007/s11390-023-1302-6","url":null,"abstract":"<p>The scarcity of bilingual parallel corpus imposes limitations on exploiting the state-of-the-art supervised translation technology. One of the research directions is employing relations among multi-modal data to enhance performance. However, the reliance on manually annotated multi-modal datasets results in a high cost of data labeling. In this paper, the topic semantics of images is proposed to alleviate the above problem. First, topic-related images can be automatically collected from the Internet by search engines. Second, topic semantics is sufficient to encode the relations between multi-modal data such as texts and images. Specifically, we propose a visual topic semantic enhanced translation (VTSE) model that utilizes topic-related images to construct a cross-lingual and cross-modal semantic space, allowing the VTSE model to simultaneously integrate the syntactic structure and semantic features. In the above process, topic similar texts and images are wrapped into groups so that the model can extract more robust topic semantics from a set of similar images and then further optimize the feature integration. The results show that our model outperforms competitive baselines by a large margin on the Multi30k and the Ambiguous COCO datasets. Our model can use external images to bring gains to translation, improving data efficiency.</p>","PeriodicalId":50222,"journal":{"name":"Journal of Computer Science and Technology","volume":"37 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139657107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}