检索结果-内蒙古大学图书馆

Iterative Refinement of Project-Level Code Context for Precise Code Generation with Compiler Feedback

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Bi, Zhangqian Wan, Yao Wang, Zheng Zhang, Hongyu Guan, Batu Lu, Fangxin Zhang, Zili Sui, Yulei Jin, Hai Shi, Xuanhua Huazhong University of Science and Technology China University of Leeds United Kingdom Chongqing University China Shanghai Jiao Tong University China University of New South Wales Australia National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China

Large Language Models (LLMs) have shown remarkable progress in automated code generation. Yet, LLM-generated code may contain errors in API usage, class, data structure, or missing project-specific information. As much of this project-specific context cannot fit into the prompts of LLMs, we must find ways to allow the model to explore the project-level code context. We present COCOGEN, a new code generation approach that uses compiler feedback to improve the LLM-generated code. COCOGEN first leverages static analysis to identify mismatches between the generated code and the project’s context. It then iteratively aligns and fixes the identified errors using information extracted from the code repository. We integrate COCOGEN with two representative LLMs, i.e., GPT-3.5-Turbo and Code Llama (13B), and apply it to Python code generation. Experimental results show that COCOGEN significantly improves the vanilla LLMs by over 80% in generating code dependent on the project context and consistently outperforms the existing retrieval-based code generation baselines. © 2024, CC BY-NC-SA.

关键词： Static analysis

Multi-stage network embedding for exploring heterogeneous edges

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Huang, Hong Song, Yu Ye, Fanghua Xie, Xing Shi, Xuanhua Jin, Hai The National Engineering Research Center for Big Data Technology Service Computing Technology and System Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China The Department of Computer Science University College London London United Kingdom Microsoft Research Asia Beijing China

The relationships between objects in a network are typically diverse and complex, leading to the heterogeneous edges with different semantic information. In this paper, we focus on exploring the heterogeneous edges for network representation learning. By considering each relationship as a view that depicts a specific type of proximity between nodes, we propose a multi-stage non-negative matrix factorization (MNMF) model, committed to utilizing abundant information in multiple views to learn robust network representations. In fact, most existing network embedding methods are closely related to implicitly factorizing the complex proximity matrix. However, the approximation error is usually quite large, since a single low-rank matrix is insufficient to capture the original information. Through a multi-stage matrix factorization process motivated by gradient boosting, our MNMF model achieves lower approximation error. Meanwhile, the multi-stage structure of MNMF gives the feasibility of designing two kinds of non-negative matrix factorization (NMF) manners to preserve network information better. The united NMF aims to preserve the consensus information between different views, and the independent NMF aims to preserve unique information of each view. Concrete experimental results on realistic datasets indicate that our model outperforms three types of baselines in practical applications. © 2021, CC BY.

关键词： Non-negative matrix factorization

Metadata-Driven Federated Learning of Connectional Brain Templates in Non-IID Multi-Domain Scenarios

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Chen, Geng Wang, Qingyue Rekik, Islem National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology School of Computer Science and Engineering Northwestern Polytechnical University China BASIRA Lab Imperial-X Department of Computing Imperial College London United Kingdom

A connectional brain template (CBT) is a holistic representation of a population of multi-view brain connectivity graphs, encoding shared patterns and normalizing typical variations across individuals. The federation of CBT learning allows for an inclusive estimation of representative center of multi-domain brain connectivity datasets in a fully data-preserving manner. However, existing methods overlook the non-independent and identically distributed (non-IDD) issue stemming from multidomain brain connectivity heterogeneity, in which data domains are drawn from different hospitals and imaging modalities. This non-IDD issue degrades the centrality of locally learned CBT from multiple decentralized datasets (i.e., domains/hospitals), eventually leading to a limited representation ability. To overcome this limitation, we unprecedentedly propose a metadata-driven federated learning framework, called MetaFedCBT, for cross-domain CBT learning. Given the data drawn from a specific domain (i.e., hospital), our model aims to learn a metadata in a fully supervised manner by introducing a local client-based regressor network. The generated meta-data is forced to meet the statistical attributes (e.g., mean) of other domains, while preserving their privacy. Our supervised meta-data generation approach boosts the unsupervised learning of a more centered and representative and holistic CBT of a particular brain state across diverse domains. As the federated learning progresses over multiple rounds, the learned metadata and associated generated connectivities are are continuously updated to better approximate the target domain information. MetaFedCBT overcomes the non-IID issue of existing methods by generating informative brain connectivities for privacy-preserving holistic CBT learning with the guidance using metadata. Extensive experiments on multi-view morphological brain networks of normal and patient subjects demonstrate that our MetaFedCBT is a superior federated CBT learnin

关键词： Metadata

Insulating materials for realising carbon neutrality:Opportunities,remaining issues and challenges

学校读者我要写书评

暂无评论

High Voltage 2022年第4期7卷 610-632页

作者： Chuanyang Li Yang Yang Guoqiang Xu Yao Zhou Mengshuo Jia Shaolong Zhong Yu Gao Chanyeop Park Qiang Liu Yalin Wang Shakeel Akram Xiaoliang Zeng Yi Li Fangwei Liang Bin Cui Junpeng Fang Lingling Tang Yulin Zeng Xingtao Hu Jiachen Gao Giovanni Mazzanti Jinliang He Jianxiao Wang Davide Fabiani Gilbert Teyssedre Yang Cao Feipeng Wang Yunlong Zi The State Key Laboratory of Power System Department of Electrical EngineeringTsinghua UniversityBeijingChina Department of Mechanical and Automation Engineering The Chinese University of Hong KongHong KongChina Simpson Querrey Institute Northwestern UniversityEvanstonIllinoisUSA Department of Materials Science and Engineering The Pennsylvania State UniversityUniversity ParkPennsylvaniaUSA Department of Information Technology and Electrical Engineering ETH ZürichZürichSwitzerland School of Electrical and Information Engineering Tianjin UniversityTianjinChina Department of Electrical and Computer Engineering Mississippi State UniversityMississippi StateMississippiUSA Electrical and Electronic Engineering The University of ManchesterManchesterUK Department of Electrical Engineering School of Electronic Information and Electrical EngineeringShanghai Jiao Tong UniversityShanghaiChina Key Laboratory of Control of Power Transmission and Conversion(SJTU) Ministry of EducationShanghaiChina College of Electrical Engineering Sichuan UniversityChengduChina Shenzhen Institute of Advanced Technology Chinese Academy of Sciences ShenzhenShenzhenChina School of Electrical Engineering Wuhan UniversityWuhanChina School of Integrated Circuits Tsinghua UniversityBeijingChina Jiangsu Jinxin Electric Appliance Co.Ltd. YangzhouChina Tai'an Tiancheng Safety Evaluation Co. Ltd.Tai'anChina Department Electrical Electronic and Information EngineeringUniversity of BolognaBolognaItaly National Engineering Laboratory for Big Data Analysis and Applications Peking UniversityBeijingChina Laplace Paul Sabatier Universityand CNRSToulouseFrance Electrical and Computer Engineering University of ConnecticutStorrsConnecticutUSA State Key Laboratory of Power Transmission Equipment&System Security and New Technology School of Electrical EngineeringChongqing UniversityChongqingChina Thrust of Sustainable Energy and Environment The Hong Kong University of Science and Technology(Guangzhou)GuangzhouChina De

The 2050 carbon-neutral vision spawns a novel energy structure revolution,and the construction of the future energy structure is based on equipment *** material,as the core of electrical power equipment and electrified transportation asset,faces unprecedented challenges and *** goal of carbon neutral and the urgent need for innovation in electric power equipment and electrification assets are first *** engineering challenges constrained by the insulation system in future electric power equipment/devices and electrified transportation assets are *** materials,including intelligent insulating material,high thermal conductivity insulating material,high energy storage density insulating material,extreme environment resistant insulating material,and environmental-friendly insulating material,are cat-egorised with their scientific issues,opportunities and challenges under the goal of carbon neutrality being *** the context of carbon neutrality,not only improves the understanding of the insulation problems from a macro level,that is,electrical power equipment and electrified transportation asset,but also offers opportunities,remaining issues and challenges from the insulating material *** is hoped that this paper en-visions the challenges regarding design and reliability of insulations in electrical equipment and electric vehicles in the context of policies towards carbon neutrality *** authors also hope that this paper can be helpful in future development and research of novel insulating materials,which promote the realisation of the carbon-neutral vision.

关键词： neutral insulation extreme

STSE-xLSTM: A Deep Learning Framework for Automated Seizure Detection in Long Video Sequences Using Spatio-Temporal and Attention Mechanisms

学校读者我要写书评

暂无评论

STSE-xLSTM: A Deep Learning Framework for Automated Seizure ...

International Conference on Computer and Communications (ICCC)

作者： Lihui Ding Hongliang Wang Lijun Fu Shenyang Institute of Computing Technology University of Chinese Academy of Sciences Shenyang China Liaoning Province Human-Computer Interaction System Engineering Research Center Based on Digital Twin Shenyang China Laboratory of Big Data and Artificial Intelligence Technology Shandong University Beijing China

ISBN: (数字)9798331507077

ISBN: (纸本)9798331507084

A detailed analysis of seizure semiology, the symptoms and signs that occur during a seizure, is crucial for the management of epilepsy patients. The inter-rater reliability of qualitative visual analysis is often low, highlighting the need for automatic and quantitative analysis of video-recorded seizures for objective assessment. To enhance evaluation accuracy and consistency, computer-aided video analysis of seizures has emerged as a natural direction of development. In the field of medical applications, deep learning and computer vision methods have driven significant advancements. However, analyzing video data presents challenges due to high dimensionality, variations in lighting, posture, and video quality. Additionally, the duration of seizures can range from seconds to minutes. To effectively extract features from seizures, we propose a novel deep learning model, STSE-xLSTM, designed to learn deep representations of arbitrarily long seizure videos. This model integrates a Spatio-Temporal Convolutional Network (STCNN) for feature extraction, a Squeeze-and-Excitation (SE) attention mechanism to enhance relevant features, and an extended Long Short-Term Memory network (xLSTM) to process long video sequences. The SE mechanism adaptively recalibrates channel weights, improving recognition of minority classes and reducing information redundancy. The xLSTM introduces scalar and matrix memory units, optimizing memory capacity and multi-scale feature extraction, thereby enhancing detection accuracy. When evaluated on the GESTURES dataset, our model outperformed state-of-the-art epilepsy detection methods in classification performance. Key contributions include the integration of STCNN, SE attention mechanism, and xLSTM to address data imbalance issues and improve the handling of complex sequential data in epilepsy detection.

关键词： Deep learning Adaptation models Visualization Attention mechanisms Accuracy Video sequences Epilepsy Feature extraction Reliability Long short term memory

Automatic recognition of depression based on audio and video:A review

学校读者我要写书评

暂无评论

World Journal of Psychiatry 2024年第2期14卷 225-233页

作者： Meng-Meng Han Xing-Yun Li Xin-Yu Yi Yun-Shao Zheng Wei-Li Xia Ya-Fei Liu Qing-Xiang Wang Shandong Mental Health Center Shandong UniversityJinan 250014Shandong ProvinceChina Key Laboratory of Computing Power Network and Information Security Ministry of EducationShandong Computer Science Center(National Supercomputer Center in Jinan)Qilu University of Technology(Shandong Academy of Sciences)Jinan 250353Shandong ProvinceChina Shandong Engineering Research Center of Big Data Applied Technology Faculty of Computer Science and TechnologyQilu University of Technology(Shandong Academy of Sciences)Jinan 250353Shandong ProvinceChina Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer ScienceJinan 250353Shandong ProvinceChina Department of Ward Two Shandong Mental Health CenterShandong UniversityJinan 250014Shandong ProvinceChina

Depression is a common mental health *** current depression detection methods,specialized physicians often engage in conversations and physiological examinations based on standardized scales as auxiliary measures for depression ***-biological markers-typically classified as verbal or non-verbal and deemed crucial evaluation criteria for depression-have not been effectively *** physicians usually require extensive training and experience to capture changes in these *** in deep learning technology have provided technical support for capturing non-biological *** researchers have proposed automatic depression estimation(ADE)systems based on sounds and videos to assist physicians in capturing these features and conducting depression *** article summarizes commonly used public datasets and recent research on audio-and video-based ADE based on three perspectives:datasets,deficiencies in existing research,and future development directions.

关键词： Depression recognition Deep learning Automatic depression estimation system Audio processing Image processing Feature fusion Future development

iGniter: Interference-Aware GPU Resource Provisioning for Predictable DNN Inference in the Cloud

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Xu, Fei Xu, Jianian Chen, Jiabin Chen, Li Shang, Ruitao Zhou, Zhi Liu, Fangming The Shanghai Key Laboratory of Multidimensional Information Processing School of Computer Science and Technology East China Normal University 3663 N. Zhongshan Road Shanghai200062 China The School of Computing and Informatics University of Louisiana at Lafayette 301 East Lewis Street LafayetteLA70504 United States The Guangdong Key Laboratory of Big Data Analysis and Processing School of Computer Science and Engineering Sun Yat-sen University 132 E. Waihuan Road Guangzhou510006 China The National Engineering Research Center for Big Data Technology and System The Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology 1037 Luoyu Road Wuhan430074 China

GPUs are essential to accelerating the latency-sensitive deep neural network (DNN) inference workloads in cloud datacenters. To fully utilize GPU resources, spatial sharing of GPUs among co-located DNN inference workloads becomes increasingly compelling. However, GPU sharing inevitably brings severe performance interference among co-located inference workloads, as motivated by an empirical measurement study of DNN inference on EC2 GPU instances. While existing works on guaranteeing inference performance service level objectives (SLOs) focus on either temporal sharing of GPUs or reactive GPU resource scaling and inference migration techniques, how to proactively mitigate such severe performance interference has received comparatively little attention. In this paper, we propose iGniter, an interference-aware GPU resource provisioning framework for cost-efficiently achieving predictable DNN inference in the cloud. iGniter is comprised of two key components: (1) a lightweight DNN inference performance model, which leverages the system and workload metrics that are practically accessible to capture the performance interference;(2) A cost-efficient GPU resource provisioning strategy that jointly optimizes the GPU resource allocation and adaptive batching based on our inference performance model, with the aim of achieving predictable performance of DNN inference workloads. We implement a prototype of iGniter based on the NVIDIA Triton inference server hosted on EC2 GPU instances. Extensive prototype experiments on four representative DNN models and datasets demonstrate that iGniter can guarantee the performance SLOs of DNN inference workloads with practically acceptable runtime overhead, while saving the monetary cost by up to 25% in comparison to the state-of-the-art GPU resource provisioning strategies. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep neural networks

PoE: a Panel of Experts for Generalized Automatic Dialogue Assessment

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhang, Chen D'Haro, Luis Fernando Zhang, Qiquan Friedrichs, Thomas Li, Haizhou The Human Language Technology Group at Electrical & Computer Engineering Department National University of Singapore Singapore Spain Pte Ltd Singapore The Guangdong Provincial Key Laboratory of Big Data Computing The Chinese University of Hong Kong Shenzhen China

Chatbots are expected to be knowledgeable across multiple domains, e.g. for daily chit-chat, exchange of information, and grounding in emotional situations. To effectively measure the quality of such conversational agents, a model-based automatic dialogue evaluation metric (ADEM) is expected to perform well across multiple domains. Despite significant progress, an ADEM that works well in one domain does not necessarily generalize to another. This calls for a dedicated network architecture for domain generalization. To tackle the multi-domain dialogue evaluation task, we propose a Panel of Experts (PoE), a multitask network that consists of a shared transformer encoder and a collection of lightweight adapters. The shared encoder captures the general knowledge of dialogues across domains, while each adapter specializes in one specific domain and serves as a domain expert. To validate the idea, we construct a high-quality multi-domain dialogue dataset leveraging data augmentation and pseudo labeling. The PoE network is comprehensively assessed on 16 dialogue evaluation datasets spanning a wide range of dialogue domains. It achieves state-of-the-art performance in terms of mean Spearman correlation over all the evaluation datasets. It exhibits better zero-shot generalization than existing state-of-the-art ADEMs and the ability to easily adapt to new domains with few-shot transfer learning. Copyright © 2022, The Authors. All rights reserved.

关键词： Network architecture

Pointsurface: Discriminative Point Cloud Surface Feature Extraction for 3d Face Recognition

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Yang, Junpeng Li, Qiufu Shen, Linlin Computer Vision Institute College of Computer Science and Software Engineering Shenzhen University Guangdong Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Guangdong Shenzhen518060 China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen University Guangdong Shenzhen518172 China

3D face recognition performs better than 2D face recognition, in terms of robustness against lighting and digital attacks. Compared to 2D data, the rich geometric information in 3D data could be very useful to improve the performance of 3D face recognition. This paper introduces PointSurFace, a surface feature extraction integrated model tailored for point cloud based 3D face recognition. PointSurFace adopts a specially designed surface feature extraction module, which explores information about the position and normals of the 3D face surface to explicitly extract local discriminative features from 3D point cloud faces. Subsequently, it alternates between set abstraction (SA) layers and Inverted Residual MLP (InvResMLP) modules to effectively extract features from the point cloud data. To better handle the spatial relationships in the point cloud data, PointSurFace integrates coordinate and feature data within each SA layer and InvResMLP module. It also applies channel fusion operations to align these different types of data, thereby more effectively extracts point cloud features. Experimental results in 3D facial recognition and verification across multiple datasets demonstrate that PointSurFace achieves the best 3D face recognition performance to date. For example, on Lock3DFace, it achieves the state-of-the-art recognition and verification accuracy of 90.03% and 80.01%, respectively, surpassing previous methods by 2.02% and 3.19%. Results on cross-quality datasets also affirm PointSurFace’s strong generalization capabilities. Ablation experiments validate the effectiveness of each module within PointSurFace, which suggests that the proposed surface feature extraction module significantly enhances the point cloud based 3D facial recognition performance. © 2024, The Authors. All rights reserved.

关键词： Feature extraction