检索结果-内蒙古大学图书馆

System Report for CCL24-Eval Task 9:Chinese Vision-Language Understanding Evaluation 23

23rd Chinese National Conference on Computational Linguistics, CCL 2024

作者： Wang, Jiangkuo Zheng, Linwei Chen, Kehai Bai, Xuefeng Zhang, Min School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

ISBN: (纸本)9780000000002

This paper introduces our systems submitted for the Chinese Vision-Language Understanding Evaluation task at the 23rd Chinese Computational Linguistics Conference. In this competition, we utilized X2-VLM and CCLM models to participate in various subtasks such as image-text retrieval, visual grounding, visual dialogue, and visual question answering. Additionally, we employed other models to assess performance on certain subtasks. We optimized our models and successfully applied them to these different tasks. © 2024 China National Conference on Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Optimal Transport-based Labor-free Text Prompt Modeling for Sketch Re-identification 38

Optimal Transport-based Labor-free Text Prompt Modeling for ...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Li, Rui Ren, Tingting Wen, Jie Li, Jinxing School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

Sketch Re-identification (Sketch Re-ID), which aims to retrieve target person from an image gallery based on a sketch query, is crucial for criminal investigation, law enforcement, and missing person searches. Existing methods aim to alleviate the modality gap by employing semantic metrics constraints or auxiliary modal guidance. However, they incur expensive labor costs and inevitably omit fine-grained modality-consistent information due to the abstraction of sketches. To address this issue, this paper proposes a novel Optimal Transport-based Labor-free Text Prompt Modeling (OLTM) network, which hierarchically extracts coarse- and fine-grained similarity representations guided by textual semantic information without any additional annotations. Specifically, multiple target attributes are flexibly obtained by a pre-trained visual question answering (VQA) model. Subsequently, a text prompt reasoning module employs learnable prompt strategy and optimal transport algorithm to extract discriminative global and local text representations, which serve as a bridge for hierarchical and multi-granularity modal alignment between sketch and image modalities. Additionally, instead of measuring the similarity of two samples by only computing their distance, a novel triplet assignment loss is further proposed, in which the whole data distribution also contributes to optimizing the inter/intra-class distances. Extensive experiments conducted on two public benchmarks consistently demonstrate the robustness and superiority of our OLTM over state-of-the-art methods. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Chain-of-Thought Enhanced Content Detection in Large Language Models 3

Chain-of-Thought Enhanced Content Detection in Large Languag...

引用

3rd International Conference on Cloud Computing, Big Data Application and Software Engineering, CBASE 2024

作者： Zou, Chenjun Feng, Zhuoyan Guan, Hongwei Xing, Chuanjun School of Computer Science and Technology Heilongjiang Institute of Technology Harbin China

ISBN: (纸本)9798331506582

This paper proposes and explores a content detection method for large language models based on the Chain of Thought (CoT) logic chain. By incrementally extracting logic chains from text or code and converting them into feature vectors, this method enables logical plagiarism detection and content verification for complex tasks. Particularly for classical algorithmic problems, such as dynamic programming, this approach effectively mitigates plagiarism attempts through variable renaming. The paper first delves into the theoretical foundation of CoT logic chains and analyzes their application mechanisms within large language models. It then provides a detailed account of the steps for feature extraction and similarity detection algorithms, including how to efficiently convert logic chains into feature vectors for comparison. To validate the effectiveness of this method, several classical algorithmic instances were selected for experimentation. Results indicate that the CoT logic chain-based content detection method offers significant advantages in detecting originality in academic papers and code. This approach not only substantially improves the detection rate of academic misconduct but also assists editors and reviewers in more accurately identifying AI-generated content, thereby holding broad applications in academic publishing, code review, and educational fields. Additionally, the paper discusses the potential application of this method in other domains, such as patent plagiarism detection and legal document similarity analysis. © 2024 IEEE.

关键词： Algorithmic languages

来源：评论

学校读者我要写书评

暂无评论

LLDif: Diffusion Models for Low-Light Facial Expression Recognition 27th

LLDif: Diffusion Models for Low-Light Facial Expression Rec...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Wang, Zhifeng Zhang, Kaihao Sankaranarayana, Ramesh College of Engineering and Computer Science Australian National University CanberraACT Australia School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

ISBN: (纸本)9783031782008

This paper introduces LLDif, a novel diffusion-based facial expression recognition (FER) framework tailored for extremely low-light (LL) environments. Images captured under such conditions often suffer from low brightness and significantly reduced contrast, presenting challenges to conventional methods. These challenges include poor image quality that can significantly reduce the accuracy of emotion recognition. LLDif addresses these issues with a novel two-stage training process that combines a Label-aware CLIP (LA-CLIP), an embedding prior network (PNET), and a transformer-based network adept at handling the noise of low-light images. The first stage involves LA-CLIP generating a joint embedding prior distribution (EPD) to guide the LLformer in label recovery. In the second stage, the diffusion model (DM) refines the EPD inference, ultilising the compactness of EPD for precise predictions. Experimental evaluations on various LL-FER datasets have shown that LLDif achieves competitive performance, underscoring its potential to enhance FER applications in challenging lighting conditions. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Practical Deterministic Transaction Processing with Low-cost Re-execution 30

Practical Deterministic Transaction Processing with Low-cost...

引用

30th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2024

作者： Li, Xingchen Wang, Xinyuan Huang, Hejiao Harbin Institute of Technology School of Computer Science and Technology Shenzhen China

ISBN: (纸本)9798331515966

Deterministic concurrency control (DCC) can guarantee consistent execution results with the same input transactions, which can reduce the synchronization overhead when applied to a multi-replica distributed database. To achieve determinism, early DCC protocols require priori knowledge about execution details of transactions, such as read/write sets, for prescheduling. State-of-the-art DCC protocols like Aria and FGA address this impractical limitation. However, they struggle with performance degradation due to high abort rate under skewed and write-intensive workloads. In this paper, we present Dmucca, a practical DCC protocol that maintains ideal performance across varying transaction workloads without requiring any priori knowledge about them. Dmucca utilizes re-execution strategy to commit nearly all transactions after their second execution. To reduce the cost of re-execution, Dmucca incorporates a multiversion cache, inter-batch pipeline parallel processing, as well as a novel reordering schema based on weight prediction. Finally we evaluate Dmucca as well as other state-of-the-art DCC protocols with the two popular benchmarks, YCSB and TPC-C. And the experimental results show that Dmucca has the optimal performance in most scenarios. © 2024 IEEE.

关键词： Concurrency control

来源：评论

学校读者我要写书评

暂无评论

Minimum Epsilon-Kernel Computation for Large-Scale Data Processing

引用

Journal of computer science & technology 2022年第6期37卷 1398-1411页

作者： Hong-Jie Guo Jian-Zhong Li Hong Gao School of Computer Science and Technology Harbin Institute of TechnologyHarbin 150001China Faculty of Computer Science and Control Engineering Shenzhen Institute of Advanced TechnologyChinese Academy of SciencesShenzhen 518000China

Kernel is a kind of data summary which is elaborately extracted from a large *** a problem,the solution obtained from the kernel is an approximate version of the solution obtained from the whole dataset with a provable approximate *** is widely used in geometric optimization,clustering,and approximate query processing,etc.,for scaling them up to massive *** this paper,we focus on the minimumε-kernel(MK)computation that asks for a kernel of the smallest size for large-scale data *** the open problem presented by Wang et *** whether the minimumε-coreset(MC)problem and the MK problem can be reduced to each other,we first formalize the MK problem and analyze its *** to the NP-hardness of the MK problem in three or higher dimensions,an approximate algorithm,namely Set Cover-Based Minimumε-Kernel algorithm(SCMK),is developed to solve *** prove that the MC problem and the MK problem can be Turing-reduced to each ***,we discuss the update of MK under insertion and deletion operations,***,a randomized algorithm,called the Randomized Algorithm of Set Cover-Based Minimumε-Kernel algorithm(RA-SCMK),is utilized to further reduce the complexity of *** efficiency and effectiveness of SCMK and RA-SCMK are verified by experimental results on real-world and synthetic *** show that the kernel sizes of SCMK are 2x and 17.6x smaller than those of an ANN-based method on real-world and synthetic datasets,*** speedup ratio of SCMK over the ANN-based method is 5.67 on synthetic ***-SCMK runs up to three times faster than SCMK on synthetic datasets.

关键词： approximate query processing kernel large-scale dataset NP-hard

来源：评论

学校读者我要写书评

暂无评论

Structured Matrix Basis for Multivariate Time Series Forecasting with Interpretable Dynamics 38

Structured Matrix Basis for Multivariate Time Series Forecas...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Chen, Xiaodan Li, Xiucheng Chen, Xinyang Li, Zhijun School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

Multivariate time series forecasting is of central importance in modern intelligent decision systems. The dynamics of multivariate time series are jointly characterized by temporal dependencies and spatial correlations. Hence, it is equally important to build the forecasting models from both perspectives. The real-world multivariate time series data often presents spatial correlations that show structures and evolve dynamically. To capture such dynamic spatial structures, the existing forecasting approaches often rely on a two-stage learning process (learning dynamic series representations and then generating spatial structures), which is sensitive to the small time-window input data and has high variance. To address this, we propose a novel forecasting model with a structured matrix basis. At its core is a dynamic spatial structure generation function whose output space is well-constrained and the generated structures have lower variance, meanwhile, it is more expressive and can offer interpretable dynamics. This is achieved via a novel structured parameterization and imposing structure regularization on the matrix basis. The resulting forecasting model can achieve up to 8.5% improvements over the existing methods on six benchmark datasets, and meanwhile, it enables us to gain insights into the dynamics of underlying systems. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Att-EMD-Unet: A Novel weakly supervised perspective for ECG segmentation

Att-EMD-Unet: A Novel weakly supervised perspective for ECG ...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Lin, Yuxin Wang, Wei Harbin Institute of Technology School of Computer Science and Technology Shenzhen China

ISBN: (纸本)9798350386226

Automatic ECG segmentation has gained significant attention due to its critical role in cardiac analysis and diagnosis. However, current automatic ECG segmentation methods are hindered by the need for labor-intensive and expert-level annotations. To alleviate the annotation burden, we explore a weakly supervised perspective for ECG QT segmentation. Specifically, we employ annotator-friendly and less expert-intensive casual annotations as supervision signals for model training. In this paper, we propose a novel model called Att-EMD-Unet, which employs U-Net as base network structure and incorporates channel/ temporal attention mechanisms to predict the QT segments from original signals. Recognizing the challenges posed by casual and incomplete labels in our weakly supervised learning framework, we have innovatively incorporated an empirical mode decomposition (EMD) based R and T peaks attentive loss function during the training phase. This function is specifically designed to address and rectify potential inaccuracies or omissions in the estimation of R and T peaks within ECGs. An expert clinician conducted an evaluation of our proposed casual annotation method. The findings from this assessment indicate that it cuts down the time needed for labeling by about 46.58% for each beat in various ECG signals, compared to the usual detailed methods. And the experimental results reveal that our Att-EMD-Unet model surpasses conventional unsupervised methods and achieves comparable performance to state-of-the-art fully supervised learning methods. Our approach effectively overcomes the limitations of extensive labeling required in fully supervised learning, presenting an efficient and accurate solution for ECG QT segmentation © 2024 IEEE.

关键词： Electrocardiography

来源：评论

学校读者我要写书评

暂无评论

A bytecode-based integrated detection and repair method for reentrancy vulnerabilities in smart contracts

IET Blockchain

引用

IET Blockchain 2024年第3期4卷 235-251页

作者： Feng, Zijun Feng, Yuming He, Hui Zhang, Weizhe Zhang, Yu School of Cyberspace Science Harbin Institute of Technology Harbin China School of Computer Science and Technology Harbin Institute of Technology Shenzhen China Department of New Networks Peng Cheng Laboratory Shenzhen China Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies Shenzhen China

The reentrancy vulnerability in smart contracts has caused significant losses in the digital currency economy. Existing solutions for detecting and repairing this vulnerability are limited in scope and lack a comprehensive framework. Additionally, there is currently a lack of guidance methods for effectively pinpointing the location of vulnerabilities. The proposed bytecode-level method addresses these challenges by incorporating a detection module, an auxiliary localization module, and a repair module. An opcode classification method is introduced using vulnerability features and a BiLSTM-Attention-based sequence model to enhance detection accuracy. To overcome difficulties in vulnerability localization, an auxiliary localization method based on data flow and control flow analysis is proposed, enabling developers to better locate vulnerabilities. Current reentrancy vulnerability repair methods are analyzed and strategies for three reachable patterns are proposed. The bytecode rewriting strategy utilizes Trampoline technology for repair, while a fuel optimization method reduces bytecode generation length to optimize gas costs. Through extensive experimental validation, the effectiveness and superiority of the proposed methods are confirmed, further validating the feasibility of the entire framework. Experimental results demonstrate that the framework offers enhanced protection against reentrancy vulnerability attacks in smart contracts. © 2023 The Authors. IET Blockchain published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and technology.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

MoME: Mixture of Multimodal Experts for Generalist Multimodal Large Language Models 38

MoME: Mixture of Multimodal Experts for Generalist Multimoda...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Shen, Leyang Chen, Gongwei Shao, Rui Guan, Weili Nie, Liqiang School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

Multimodal large language models (MLLMs) have demonstrated impressive capabilities across various vision-language tasks. However, a generalist MLLM typically underperforms compared with a specialist MLLM on most VL tasks, which can be attributed to task interference. In this paper, we propose a mixture of multimodal experts (MoME) to mitigate task interference and obtain a generalist MLLM. Our MoME is composed of two key components, a mixture of vision experts (MoVE) and a mixture of language experts (MoLE). MoVE can adaptively modulate the features transformed from various vision encoders, and has a strong compatibility in transformation architecture. MoLE incorporates sparsely gated experts into LLMs to achieve painless improvements with roughly unchanged inference costs. In response to task interference, our MoME specializes in both vision and language modality to adapt to task discrepancies. Extensive experiments show that MoME significantly improves the performance of generalist MLLMs across various VL tasks. © 2024 Neural information processing systems foundation. All rights reserved.

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：