检索结果-内蒙古大学图书馆

6th International Conference on Intelligent Computing and Signal Processing (ICSP)

作者： Liquan Wang Shu Zhan School of Computer and Information Engineering Hefei University of Technology Key Laboratory of Big Data Knowledge Engineering Ministry of Education Hefei China

Micro Expression (ME) is the subtle facial expressions that people show when they express their inner feelings. To address the problem that micro-expression recognition is difficult and less accurate due to the small number of samples and uneven distribution of different categories, we propose a model framework to improve the accuracy of micro-expression recognition. The peak frames containing more key expression information in the micro-expression video sequences are extracted; SE-ResNeXt-50, an improved residual network with SE module, is used to extract features from the peak frames of micro-expressions, where the SE module can better learn the key information in the features, and ResNeXt simplifies the structure by replacing the dense structure with the sparse structure through group convolution, which improves the recognition efficiency. The recognition efficiency is improved by replacing the dense structure with the sparse structure by group convolution. At the same time, the Focal Loss loss function can better solve the model performance problem caused by the imbalance of micro-expression data. Simulation experiments are conducted on the micro-expression dataset CASMEⅡ, and it is found that the improved residual network and peak frame improve the accuracy and F1 value of micro-expression recognition. The improved residual network and peak frame can reduce the effect of small data set, make the model have good fitting effect, and improve the performance of different categories, improve the recognition accuracy of micro-expressions, and have better recognition performance for micro-expression recognition.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Rethinking data-Free Quantization as a Zero-Sum Game

arXiv

引用

arXiv 2023年

作者： Qian, Biao Wang, Yang Hong, Richang Wang, Meng Key Laboratory of Knowledge Engineering with Big Data Ministry of Education School of Computer Science and Information Engineering Hefei University of Technology China

data-free quantization (DFQ) recovers the performance of quantized network (Q) without accessing the real data, but generates the fake sample via a generator (G) by learning from full-precision network (P) instead. However, such sample generation process is totally independent of Q, specialized as failing to consider the adaptability of the generated samples, i.e., beneficial or adversarial, over the learning process of Q, resulting into non-ignorable performance loss. Building on this, several crucial questions — how to measure and exploit the sample adaptability to Q under varied bit-width scenarios? how to generate the samples with desirable adaptability to benefit the quantized network? — impel us to revisit DFQ. In this paper, we answer the above questions from a game-theory perspective to specialize DFQ as a zero-sum game between two players — a generator and a quantized network, and further propose an Adaptability-aware Sample Generation (AdaSG) method. Technically, AdaSG reformulates DFQ as a dynamic maximization-vs-minimization game process anchored on the sample adaptability. The maximization process aims to generate the sample with desirable adaptability, such sample adaptability is further reduced by the minimization process after calibrating Q for performance recovery. The Balance Gap is defined to guide the stationarity of the game process to maximally benefit Q. The theoretical analysis and empirical studies verify the superiority of AdaSG over the state-of-the-arts. Our code is available at https://***/hfutqian/AdaSG. Copyright © 2023, The Authors. All rights reserved.

关键词： Game theory

来源：评论

学校读者我要写书评

暂无评论

Adaptive data-Free Quantization

arXiv

引用

arXiv 2023年

data-free quantization (DFQ) recovers the performance of quantized network (Q) without accessing the original data, but generates the fake sample via a generator (G) by learning from full-precision network (P), which, however, is totally independent of Q, overlooking the adaptability of the knowledge from generated samples, i.e., informative or not to the learning process of Q, resulting into the overflow of generalization error. Building on this, several critical questions - how to measure the sample adaptability to Q under varied bit-width scenarios? how to generate the samples with large adaptability to improve Q's generalization? whether the largest adaptability is the best? To answer the above questions, in this paper, we propose an Adaptive data-Free Quantization (AdaDFQ) method, which revisits DFQ from a zero-sum game perspective upon the sample adaptability between two players - a generator and a quantized network. Following this viewpoint, we further define the disagreement and agreement samples to form two boundaries, where the margin is optimized to address the over-and-under fitting issues, so as to generate the samples with adaptive adaptability to Q. Our AdaDFQ reveals: 1) the largest adaptability is NOT the best for sample generation to benefit Q's generalization;2) the knowledge of the generated sample should not be informative to Q only, but also related to the category and distribution information of the training data for P. The theoretical and empirical analysis validate the advantages of AdaDFQ over the state-of-the-arts. Our code is available at https://***/hfutqian/AdaDFQ. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Superclass Learning with Representation Enhancement

Superclass Learning with Representation Enhancement

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zeyu Gan Suyun Zhao Jinlong Kang Liyuan Shang Hong Chen Cuiping Li Key Lab of Data Engineering and Knowledge Engineering of MOE Renmin University of China Beijing China Renmin University of China Beijing China

In many real scenarios, data are often divided into a handful of artificial super categories in terms of expert knowledge rather than the representations of images. Concretely, a superclass may contain massive and various raw categories, such as refuse sorting. Due to the lack of common semantic features, the existing classification techniques are intractable to recognize superclass without raw class labels, thus they suffer severe performance damage or require huge annotation costs. To narrow this gap, this paper proposes a superclass learning framework, called SuperClass Learning with Representation Enhancement(SCLRE), to recognize super categories by leveraging enhanced representation. Specifically, by exploiting the self-attention technique across the batch, SCLRE collapses the boundaries of those raw categories and enhances the representation of each superclass. On the enhanced representation space, a superclassaware decision boundary is then reconstructed. Theoretically, we prove that by leveraging attention techniques the generalization error of SCLRE can be bounded under superclass scenarios. Experimentally, extensive results demonstrate that SCLRE outperforms the baseline and other contrastive-based methods on CIFAR-100 datasets and four high-resolution datasets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

基于数据挖掘技术的税务风险检测方法评述

引用

engineering 2024年第3期34卷 43-59页

作者： Qinghua Zheng Yiming Xu Huixiang Liu Bin Shi Jiaxiang Wang Bo Dong School of Computer Science and Technology Xi’an Jiaotong UniversityXi’an 710049China Shaanxi Provincial Key Laboratory of Big Data Knowledge Engineering Xi’an Jiaotong UniversityXi’an 710049China School of Distance Education Xi’an Jiaotong UniversityXi’an 710049China

Tax risk behavior causes serious loss of fiscal revenue,damages the country’s public infrastructure,and disturbs the market economic order of fair *** recent years,tax risk detection,driven by information technology such as data mining and artificial intelligence,has received extensive *** promote the high-quality development of tax risk detection methods,this paper provides the first comprehensive overview and summary of existing tax risk detection methods *** specifi-cally,it first discusses the causes and negative impacts of tax risk behaviors,along with the development of tax risk *** then focuses on data-mining-based tax risk detection methods utilized around the *** on the different principles employed by the algorithms,existing risk detection methods can be divided into two categories:relationship-based and non-relationship-based.A total of 14 risk detection methods are identified,and each method is thoroughly explored and ***,four major technical bottlenecks of current data-driven tax risk detection methods are analyzed and discussed,including the difficulty of integrating and using fiscal and tax fragmented knowledge,unexplainable risk detection results,the high cost of risk detection algorithms,and the reliance of existing algorithms on labeled *** investigating these issues,it is concluded that knowledge-guided and datadriven big data knowledge engineering will be the development trend in the field of tax risk in the future;that is,the gradual transition of tax risk detection from informatization to intelligence is the future development direction.

关键词： Tax risk detection data mining knowledge guide Informatization Intellectualization

来源：评论

学校读者我要写书评

暂无评论

EdgeNN: Efficient Neural Network Inference for CPU-GPU Integrated Edge Devices

EdgeNN: Efficient Neural Network Inference for CPU-GPU Integ...

引用

International Conference on data engineering

作者： Chenyang Zhang Feng Zhang Kuangyu Chen Mingjun Chen Bingsheng He Xiaoyong Du Key Laboratory of Data Engineering and Knowledge Engineering (MOE) and School of Information Renmin University of China School of Computing National University of Singapore

With the development of the architectures and the growth of AIoT application requirements, data processing on edge has become popular. Neural network inference is widely employed for data analytics on edge devices. This paper extensively explores neural network inference on integrated edge devices and proposes EdgeNN, the first neural network inference solution on CPU-GPU integrated edge devices. EdgeNN has three novel characteristics. First, EdgeNN can adaptively utilize the unified physical memory and conduct the zero-copy optimization. Second, EdgeNN involves a novel inference-targeted inter- and intra-kernel CPU-GPU hybrid execution approach, which co-runs the CPU with the GPU to fully utilize the edge device’s computing resources. Third, EdgeNN adopts a fine-grained adaptive inference tuning approach, which can divide the complicated inference structure into sub-tasks mapped to the CPU and the GPU. Experiments show that on six popular neural network inference tasks, EdgeNN brings an average of 3.97×, 3.12×, and 8.80× speedups to inference on the CPU of the integrated device, inference on a mobile phone CPU, and inference on an edge CPU device. Additionally, it achieves 22.02% time benefits to the direct execution of the original programs. Specifically, 9.93% comes from better utilization of unified memory, and 10.76% comes from CPU-GPU hybrid execution. Besides, EdgeNN can deliver 29.14× and 5.70× higher energy efficiency than the edge CPU and the discrete GPU, respectively. We have made EdgeNN available at https://***/ChenyangZhang-cs/EdgeNN.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Generative-Contrastive Graph Learning for Recommendation

arXiv

引用

arXiv 2023年

作者： Yang, Yonghui Wu, Zhengwei Wu, Le Zhang, Kun Hong, Richang Zhang, Zhiqiang Zhou, Jun Wang, Meng Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology China Ant Group China Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology Institute of Artificial Intelligence Hefei Comprehensive National Science Center China

By treating users’ interactions as a user-item graph, graph learning models have been widely deployed in Collaborative Filtering (CF) based recommendation. Recently, researchers have introduced Graph Contrastive Learning (GCL) techniques into CF to alleviate the sparse supervision issue, which first constructs contrastive views by data augmentations and then provides self-supervised signals by maximizing the mutual information between contrastive views. Despite the effectiveness, we argue that current GCL-based recommendation models are still limited as current data augmentation techniques, either structure augmentation or feature augmentation. First, structure augmentation randomly dropout nodes or edges, which is easy to destroy the intrinsic nature of the user-item graph. Second, feature augmentation imposes the same scale noise augmentation on each node, which neglects the unique characteristics of nodes on the graph. To tackle the above limitations, we propose a novel Variational Graph Generative-Contrastive Learning (VGCL) framework for recommendation. Specifically, we leverage variational graph reconstruction to estimate a Gaussian distribution of each node, then generate multiple contrastive views through multiple samplings from the estimated distributions, which builds a bridge between generative and contrastive learning. The generated contrastive views can well reconstruct the input graph without information distortion. Besides, the estimated variances are tailored to each node, which regulates the scale of contrastive loss for each node on optimization. Considering the similarity of the estimated distributions, we propose a cluster-aware twofold contrastive learning, a node-level to encourage consistency of a node’s contrastive views and a cluster-level to encourage consistency of nodes in a cluster. Finally, extensive experimental results on three public datasets clearly demonstrate the effectiveness of the proposed model. Copyright © 2023, The Authors.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

An Actor-centric Causality Graph for Asynchronous Temporal Inference in Group Activity

An Actor-centric Causality Graph for Asynchronous Temporal I...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zhao Xie Tian Gao Kewei Wu Jiao Chang Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology School of Computer Science and Information Engineering Hefei University of Technology

The causality relation modeling remains a challenging task for group activity recognition. The causality relations describe the influence on the centric actor (effect actor) from its correlative actors (cause actors). Most existing graph models focus on learning the actor relation with synchronous temporal features, which is insufficient to deal with the causality relation with asynchronous temporal features. In this paper, we propose an Actor-Centric Causality Graph Model, which learns the asynchronous temporal causality relation with three modules, i.e., an asynchronous temporal causality relation detection module, a causality feature fusion module, and a causality relation graph inference module. First, given a centric actor and its correlative actor, we analyze their influences to detect causality relation. We estimate the self influence of the centric actor with self regression. We estimate the correlative influence from the correlative actor to the centric actor with correlative regression, which uses asynchronous features at different timestamps. Second, we synchronize the two action features by estimating the temporal delay between the cause action and the effect action. The synchronized features are used to enhance the feature of the effect action with a channel-wise fusion. Third, we describe the nodes (actors) with causality features and learn the edges by fusing the causality relation with the appearance relation and distance relation. The causality relation graph inference provides crucial features of effect action, which are complementary to the base model using synchronous relation inference. Experiments show that our method achieves state-of-the-art performance on the Volleyball dataset and Collective Activity dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

WATuning:A Workload-Aware Tuning System with Attention-Based Deep Reinforcement Learning

引用

Journal of Computer Science & Technology 2021年第4期36卷 741-761页

作者： Jia-Ke Ge Yan-Feng Chai Yun-Peng Chai Key Laboratory of Data Engineering and Knowledge Engineering of Ministry of Education Renmin University of China Beijing 100872China School of Information Renmin University of ChinaBeijing 100872China College of Computer Science and Technology Taiyuan University of Science and TechnologyTaiyuan 030027China

Configuration tuning is essential to optimize the performance of systems(e.g.,databases,key-value stores).High performance usually indicates high throughput and low *** present,most of the tuning tasks of systems are performed artificially(e.g.,by database administrators),but it is hard for them to achieve high performance through tuning in various types of systems and in various *** recent years,there have been some studies on tuning traditional database systems,but all these methods have some *** this article,we put forward a tuning system based on attention-based deep reinforcement learning named WATuning,which can adapt to the changes of workload characteristics and optimize the system performance efficiently and ***,we design the core algorithm named ATT-Tune for WATuning to achieve the tuning task of *** algorithm uses workload characteristics to generate a weight matrix and acts on the internal metrics of systems,and then ATT-Tune uses the internal metrics with weight values assigned to select the appropriate ***,WATuning can generate multiple instance models according to the change of the workload so that it can complete targeted recommendation services for different types of ***,WATuning can also dynamically fine-tune itself according to the constantly changing workload in practical applications so that it can better fit to the actual environment to make *** experimental results show that the throughput and the latency of WATuning are improved by 52.6%and decreased by 31%,respectively,compared with the throughput and the latency of CDBTune which is an existing optimal tuning method.

关键词： attention mechanism auto-tuning system reinforcement learning(RL) workload-aware

来源：评论

学校读者我要写书评

暂无评论

Adaptive data-Free Quantization

Adaptive Data-Free Quantization

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Biao Qian Yang Wang Richang Hong Meng Wang Key Laboratory of Knowledge Engineering with Big Data Ministry of Education School of Computer Science and Information Engineering Hefei University of Technology China

data-free quantization (DFQ) recovers the performance of quantized network (Q) without the original data, but generates the fake sample via a generator (G) by learning from full-precision network (P), which, however, is totally independent of Q, overlooking the adaptability of the knowledge from generated samples, i.e., informative or not to the learning process of Q, resulting into the overflow of generalization error. Building on this, several critical questions — how to measure the sample adaptability to Q under varied bit-width scenarios? whether the largest adaptability is the best? how to generate the samples with adaptive adaptability to improve Q's generalization? To answer the above questions, in this paper, we propose an Adaptive data-Free Quantization (AdaDFQ) method, which revisits DFQ from a zero-sum game perspective upon the sample adaptability between two players — a generator and a quantized network. Following this viewpoint, we further define the disagreement and agreement samples to form two boundaries, where the margin between two boundaries is optimized to adaptively regulate the adaptability of generated samples to Q, so as to address the over-and-under fitting issues. Our AdaDFQ reveals: 1) the largest adaptability is NOT the best for sample generation to benefit Q's generalization; 2) the knowledge of the generated sample should not be informative to Q only, but also related to the category and distribution information of the training data for P. The theoretical and empirical analysis validate the advantages of AdaDFQ over the state-of-the-arts. Our code is available at https://***/hfutqian/AdaDFQ.

关键词：

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：