检索结果-内蒙古大学图书馆

Customizing a teacher for feature distillation

INFORMATION SCIENCES 2023年第1期640卷

作者： Tan, Chao Liu, Jie Natl Univ Def Technol Sci & Parallel & Distributed Proc Lab Changsha 410073 Peoples R China Natl Univ Def Technol Lab Software Engn Complex Syst Changsha 410073 Peoples R China

Knowledge distillation is a method to train a lightweight network by transferring class probability knowledge from a cumbersome teacher network. However, transferring only the class probability knowledge would limit the distillation performance. Therefore, several approaches have been proposed to transfer the teacher's knowledge at the feature map level. In this paper, we revisit the feature distillation method and have found that the larger the teacher's architecture/capacity becomes, the more difficult it is for the student to imitate. Thus, the feature distillation method is unable to achieve its full potential. To address this, a novel end-to-end distillation framework, termed Customizing a Teacher for Feature Distillation (CTFD), is proposed to train a teacher to be more compatible with its student. In addition, we apply the customized teacher to three feature distillation methods. Moreover, data augmentation is used as a trick to train the student to improve its generalization performance. Extensive empirical experiments and analyses are conducted on three computer vision tasks, including image classification, transfer learning, and object detection, to substantiate the effectiveness of the proposed method.

关键词： Neural network compression Knowledge distillation Knowledge transfer

来源：评论

学校读者我要写书评

暂无评论

Time Efficient Tag Searching in Large-Scale RFID Systems: A Compact Exclusive Validation Method

引用

IEEE TRANSACTIONS ON MOBILE COMPUTING 2022年第4期21卷 1476-1491页

作者： Liu, Xuan Yin, Jiangjin Liu, Jia Zhang, Shigeng Xiao, Bin Hunan Univ Coll Comp Sci & Elect Engn Changsha 410082 Peoples R China Sci & Technol Parallel & Distributed Proc Laborat Changsha 410073 Peoples R China Nanjing Univ State Key Lab Novel Software Technol Nanjing 210093 Peoples R China Cent South Univ Sch Comp Sci & Engn Changsha 410083 Peoples R China Hong Kong Polytech Univ Dept Comp Hong Kong Peoples R China

RFID technology has been widely applied in a range of applications such as inventory control, warehouse management and supply chain logistics. Many practical applications need to search a given set of tags (called wanted tags) to determine which of them are present in the system, which is usually called tag searching. Existing tag searching protocols suffer performance bottleneck and the time efficiency and need to be improved. The bottleneck stems from two factors. First, the existing methods validate the wanted tags in a random way due to the randomness of the hash function, which unavoidably generate many useless slots. Second, in order to achieve the predefined reliability requirement, the existing methods have to repeatedly validate target tags multiple times. In this paper, we design new tag searching techniques of compact exclusive validation that break through the existing performance bottleneck from two aspects. First, our protocols avoid slot waste. Different from random tag validation, our protocols validate tags orderly by mapping the wanted tags and the slots in a one-to-one manner, which makes the reader be able to validate tags in every slot. Second, our protocols avoid repeated tag responding. By combining two lightweight indicators, we rapidly filter out non-wanted tags so that there is no interference when validating the wanted tags. Hence, we need to validate each wanted tag only once, which avoids redundant validation and greatly improves time efficiency. Our protocols work with the assumption that the rough number of tags in the system can be obtained by using existing estimation algorithms, but they do not need to know exactly which tags are in the system. Theoretical analysis illustrates our methods achieve linear time complexity. Extensive experimental results show that, our best protocol can improve the time efficiency by up to 81 percent when compared with the-state-of-art solution.

关键词： Protocols Radiofrequency identification Interference Servers Cats Mobile computing Reliability Time efficiency tag searching radio frequency identification (RFID) systems

来源：评论

学校读者我要写书评

暂无评论

Efficient and Available In-Memory KV-Store with Hybrid Erasure Coding and Replication

引用

ACM TRANSACTIONS ON STORAGE 2017年第3期13卷 25-25页

作者： Chen, Haibo Zhang, Heng Dong, Mingkai Wang, Zhaoguo Xia, Yubin Guan, Haibing Zang, Binyu Shanghai Jiao Tong Univ Inst Parallel & Distributed Syst Shanghai Key Lab Scalable Comp & Syst Room 3402Software Bldg800 Dongchuan Rd Shanghai 200240 Peoples R China

In-memory key/value store (KV-store) is a key building block for many systems like databases and large websites. Two key requirements for such systems are efficiency and availability, which demand a KV-store to continuously handle millions of requests per second. A common approach to availability is using replication, such as primary-backup (PBR), which, however, requires M + 1 times memory to tolerate M failures. This renders scarce memory unable to handle useful user jobs. This article makes the first case of building highly available in-memory KV-store by integrating erasure coding to achieve memory efficiency, while not notably degrading performance. A main challenge is that an in-memory KV-store has much scattered metadata. A single KV put may cause excessive coding operations and parity updates due to excessive small updates to metadata. Our approach, namely Cocytus, addresses this challenge by using a hybrid scheme that leverages PBR for small-sized and scattered data (e.g., metadata and key), while only applying erasure coding to relatively large data (e.g., value). To mitigate well-known issues like lengthy recovery of erasure coding, Cocytus uses an online recovery scheme by leveraging the replicated metadata information to continuously serve KV requests. To further demonstrate the usefulness of Cocytus, we have built a transaction layer by using Cocytus as a fast and reliable storage layer to store database records and transaction logs. We have integrated the design of Cocytus to Memcached and extend it to support inmemory transactions. Evaluation using YCSB with different KV configurations shows that Cocytus incurs low overhead for latency and throughput, can tolerate node failures with fast online recovery, while saving 33% to 46% memory compared to PBR when tolerating two failures. A further evaluation using the SmallBank OLTP benchmark shows that in-memory transactions can run atop Cocytus with high throughput, low latency, and low abort rate and reco

关键词： Primary-backup replication erasure coding KV-store transactions

来源：评论

学校读者我要写书评

暂无评论

RF-Siamese: Approaching Accurate RFID Gesture Recognition With One Sample

引用

IEEE TRANSACTIONS ON MOBILE COMPUTING 2024年第1期23卷 797-811页

作者： Ma, Zijing Zhang, Shigeng Liu, Jia Liu, Xuan Wang, Weiping Wang, Jianxin Guo, Song Cent South Univ Sch Comp Sci & Engn Changsha 410017 Peoples R China Cent South Univ Sch Comp Sci & Engn Sci & Technol Parallel & Distributed Proc Lab PDL Changsha 410017 Peoples R China Nanjing Univ State Key Lab Novel Software Technol Nanjing 210093 Peoples R China Hunan Univ Coll Comp Sci & Elect Engn Changsha 410082 Peoples R China Hong Kong Polytech Univ Dept Comp Hong Kong Peoples R China

Performing accurate sensing in diverse environments is a challenging issue in wireless sensing technologies. Existing solutions usually require collecting a large number of samples to train a classifier for every environment, or further assume similar sample distribution between different environments such that a model trained in one environment can be transferred to another. In this paper, we propose RF-Siamese, an RFID-based gesture sensing approach that achieves comparable accuracy to existing solutions but requires only a few samples in each eivironment. RF-Siamese leverages Siamese networks to distinguish different gestures with only a small number of samples and is enhanced by several novel designs to achieve high accuracy in diverse environments. First, the network structure and parameters (e.g., loss function and distance metric) are carefully designed to be suitable for RFID gesture recognition. Second, a permutation-based dataset generation strategy is proposed to make full use of the collected samples to enhance the recognition accuracy. Third, a template matching method is proposed to extend the Siamese network to classify multiple gestures. Extensive experiments on commercial RFID devices demonstrate that RF-Siamese achieves a high accuracy of 0.93 with only one sample of each gesture when recognizing 18 different gestures, while state-of-the-art approaches based on transfer learning and meta learning achieve an accuracy of only 0.59 and 0.70, respectively.

关键词： Sensors Gesture recognition Radiofrequency identification Wireless sensor networks Wireless communication Transfer learning Mobile computing RFID few-shot learning siamese network

来源：评论

学校读者我要写书评

暂无评论

DeepEnhancer: Temporally Consistent Focal Transformer for Comprehensive Video Enhancement 24

DeepEnhancer: Temporally Consistent Focal Transformer for Co...

引用

4th Annual International Conference on Multimedia Retrieval (ICMR)

作者： Jiang, Qin Wang, Qinglin Chi, Lihua Ma, Wentao Li, Feng Liu, Jie Natl Univ Def Technol Lab Digitizing Software Frontier Equipment Sci & Technol Parallel & Distributed Proc Lab Changsha Peoples R China Hunan GuoKe Computil Technol Co Ltd Changsha Peoples R China Anhui Agr Univ Sch Informat & Artificial Intelligence Hefei Peoples R China Hefei Univ Technol Hefei Peoples R China

ISBN: (纸本)9798400706028

Restoring and colorizing old films is a comprehensive video enhancement task, marked by the presence of heterogeneous and structured degradations. Our DeepEnhancer addresses this challenge through a unified pipeline that combines restoration and colorization. In this workflow, we implement a bidirectional propagation strategy. Specifically, we incorporate second-order feature alignment to reduce the accumulation of inaccuracies in optical flow estimation. Simultaneously, we utilize cross-scale long-term attention mechanisms to model correlations within hidden states, thereby ensuring spatial and temporal consistency. To address the notable content loss in aging films, we introduce a temporally consistent focal transformer guided by global information. This transformer utilizes various window levels with distinct sub-window sizes to seamlessly integrate fine-grained and coarse-grained features. Comprehensive experimental results conclusively demonstrate the superiority of our model in both quantitative and qualitative comparisons when compared to existing approaches.

关键词： video restoration colorization Focal Transformer temporal consistency

来源：评论

学校读者我要写书评

暂无评论

Subtraction of Hyperledger Fabric: A blockchain-based lightweight storage mechanism for digital evidences

引用

JOURNAL OF SYSTEMS ARCHITECTURE 2024年 153卷

作者： Fu, Xiang Ma, Haoliang Ding, Bo Wang, Huaimin Shi, Peichang Natl Univ Def Technol Key Lab Parallel & Distributed Comp Changsha 410073 Hunan Province Peoples R China State Key Lab Complex & Crit Software Environm Changsha 410073 Hunan Province Peoples R China

After more than ten years of development, the application of blockchain has expanded beyond the financial field to include other sectors such as E-government and the medical field. One of the most renowned blockchain platforms is Hyperledger Fabric, which is widely adopted by existing works for managing digital evidences (DEs). However, these works overlook the distinct differences between DEs and token-transfer transactions and often adopt the full implementation of Hyperledger Fabric, leveraging chaincode to manage the hash of DEs while keeping the original data off-chain to conserve storage resources. This approach poses challenges such as off-chain data loss, reduced data verification efficiency, and increased resource consumption. In this paper, we introduce a lightweight storage mechanism called Blockchain Zhengju Keeper (BZK) that addresses these issues. In BZK, we subtract Hyperledger Fabric by (1) modifying the endorsement policy to reduce the risk of off-chain data loss, (2) simplifying the DE upload-onchain process while keeping the parallel execution, and (3) abandoning chaincode for DE verification to reduce latency. Comprehensive simulation experiments demonstrate the practicality of the proposed mechanism. Furthermore, compared to Hyperledger Fabric, BZK can reduce storage resource consumption by 68% and lower verification latency by 97%.

关键词： Blockchain Hyperledger fabric Lightweight storage Digital evidence

来源：评论

学校读者我要写书评

暂无评论

Eunomia: Scaling Concurrent Search Trees under Contention Using HTM 17

Eunomia: Scaling Concurrent Search Trees under Contention Us...

引用

22nd ACM SIGPLAN Symposium on Principles and Practice of parallel Programming (PPoPP)

作者： Wang, Xin Zhang, Weihua Wang, Zhaoguo Wei, Ziyun Chen, Haibo Zhao, Wenyun Fudan Univ Software Sch Shanghai Peoples R China Fudan Univ Shanghai Key Lab Data Sci Shanghai Peoples R China Fudan Univ Sch Comp Sci Shanghai Peoples R China Shanghai Jiao Tong Univ Inst Parallel & Distributed Syst Shanghai Peoples R China NYU Comp Sci Dept New York NY 10003 USA

ISBN: (纸本)9781450344937

While hardware transactional memory (HTM) has recently been adopted to construct efficient concurrent search tree structures, such designs fail to deliver scalable performance under contention. In this paper, we first conduct a detailed analysis on an HTM-based concurrent B+Tree, which uncovers several reasons for excessive HTM aborts induced by both false and true conflicts under contention. Based on the analysis, we advocate Eunomia, a design pattern for search trees which contains several principles to reduce HTM aborts, including splitting HTM regions with version based concurrency control to reduce HTM working sets, partitioned data layout to reduce false conflicts, proactively detecting and avoiding true conflicts, and adaptive con currency control. To validate their effectiveness, we apply such designs to construct a scalable concurrent B+Tree using HTM. Evaluation using key-value store benchmarks on a 20-core HTM-capable multi-core machine shows that Eunomia leads to 5X-11X speedup under high contention, while incurring small overhead under low contention.

关键词： Hardware Transactional Memory Concurrent Search Tree Opportunistic Consistency

来源：评论

学校读者我要写书评

暂无评论

You Ought to Look Around: Precise, Large Span Action Detection

You Ought to Look Around: Precise, Large Span Action Detecti...

引用

International Conference on Pattern Recognition

作者： Ge Pan Han Zhang Fan Yu Yonghong Song Yuanlin Zhang Han Yuan School of Software Engineering Xi'an Jiaotong University Xi'an China Distributed and parallel software lab Huawei Technologies Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an China

For the action localization task, pre-defined action anchors are the cornerstone of mainstream techniques. State-of-the-art models mostly rely on a dense segmenting scheme, where anchors are sampled uniformly over the temporal domain with a predefined set of scales. However, it is not sufficient because action duration varies greatly. Therefore, it is necessary for the anchors or proposals to have a variable receptive field. In this paper, we propose a method called YOLA (You Ought to Look Around) which includes three parts: 1) a robust backbone SPN-I3D for extracting spatio-temporal features. In this part, we employ a stronger backbone I3D with SPN (Segment Pyramid Network) instead of C3D to obtain multi-scale features; 2) a simple but useful feature fusion module named LFE (Local Feature Extraction). Compared with the fully connected layer and global average pooling, our LFE model is more advantageous for network to fit and fuse features. 3) a new feature segment aligning method called TPGC (Two Pathway Graph Convolution), which allows one proposal to leverage semantic features of adjacent proposals to update its content and make sure the proposals have a variable receptive field. YOLA add only a small overhead to the baseline network, and is easy to train in an end-to-end manner, running at a speed of 1097 fps. YOLA achieves a mAP of 58.3%, outperforming all existing models including both RGB-based and two stream on THUMOS'14, and achieves competitive results on ActivityNet 1.3.

关键词： Location awareness Fuses Convolution Semantics Feature extraction Pattern recognition Proposals

来源：评论

学校读者我要写书评

暂无评论

Enhancing Graph Convolutional Networks by Topology Sampling

Enhancing Graph Convolutional Networks by Topology Sampling

引用

2022 IEEE International Conference on Big Data, Big Data 2022

作者： Petit, Quentin R. Li, Chong Petiton, Serge G. Chai, Kelun Emad, Nahid Huawei Paris Research Center Université Paris-Saclay Boulogne-Billancourt France Huawei Paris Research Center Distributed and Parallel Software Lab Boulogne-Billancourt France Université de Lille Lille France Université Versailles Saint-Quentin Maison de la Simulation & LI-PaRAD Saclay France

ISBN: (纸本)9781665480451

Graph Neural Networks (GNNs) play a very important role today. It does analyze not only the graph data itself, but also the data connectivity of the graph. The quality of a GNN is thus altered by the result of extracted graph structure information. The extraction could be enhanced by GNN model design or directly from the training dataset with a GNN-decoupled method. In this paper, we propose RankedDrop, a new sampling method to improve the extraction of graph structure information. This approach is based on droppingout technique, and it adopts a spatial-aware selection of edges to drop. It takes into account structure information of the graph to control the dropping-out, and its random selection of edges to be dropped is under the control of a probability generated with respect to graph's topological importance. Our experiments point out that RankedDrop provides high-quality and robust training results compared to the leading solutions. Furthermore, RankedDrop could be a framework plugin and combined with GNN model improvements to maximize GNN quality. Furthermore, RankedDrop could be a plugin for AI frameworks like MindSpore and combined with GNN model improvements to maximize GNN quality. © 2022 IEEE.

关键词： Network topology

来源：评论

学校读者我要写书评

暂无评论

An Open Source Graph-Based Weighted Cycle Closure Method for Relative Binding Free Energy Calculations

引用

JOURNAL OF CHEMICAL INFORMATION AND MODELING 2022年第2期63卷 561-570页

作者： Li, Yishui Liu, Runduo Liu, Jie Luo, Haibin Wu, Chengkun Li, Zhe Natl Univ Def Technol Sci & Technol Parallel & Distributed Proc Lab Changsha 410073 Hunan Peoples R China Natl Univ Def Technol Lab Software Engn Complex Syst Changsha 410073 Hunan Peoples R China Sun Yat Sen Univ Sch Pharmaceut Sci Guangzhou 510275 Guangdong Peoples R China Hainan Univ Sch Pharmaceut Sci Key Lab Trop Biol Resources Minist Educ Haikou 570228 Hainan Peoples R China Natl Univ Def Technol State Key Lab High Performance Comp Changsha 410073 Hunan Peoples R China

Free energy perturbation-relative binding free energy (FEP-RBFE) prediction has shown its reliability and accuracy in the prediction of protein-ligand binding affinities, which plays a fundamental role in structure-based drug design. In FEP-RBFE predictions, the calculation of each mutation path is associated with a statistical error, and cycle closure (cc) has proven to be an effective method in improving the calculation accuracy by correcting the hysteresis (summation of errors) of each closed cycle to the theoretical value 0. However, a primary hypothesis was made in the current cycle closure method that the hysteresis is evenly distributed to all paths, which is unlikely to be true in practice and may limit the further improvement of the calculation accuracy when better error estimation methods are available. Moreover, being a closed source software makes the current cycle closure method unachievable in many studies. In this paper, a newly implemented open source graph-based weighted cycle closure (wcc) algorithm was developed and introduced, not only including functions from the original cc method but also containing a new wcc method which can consider different error contributions from different paths and further improve the calculation accuracy. The wcc program also provides a new path-independent molecular error calculation method, which can be quite useful in many studies (like structure-activity relationship (SAR)) compared with the path-dependent method of the original cc program.

关键词： Chemical calculations Free energy Genetics Ligands Molecules

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：