检索结果-内蒙古大学图书馆

KnowCoder-X: Boosting Multilingual Information Extraction via Code

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zuo, Yuxin Jiang, Wenxuan Liu, Wenxuan Li, Zixuan Bai, Long Wang, Hanbin Zeng, Yutao Jin, Xiaolong Guo, Jiafeng Cheng, Xueqi Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China School of Software Northeastern University United States School of Software Peking University China

Empirical evidence indicates that LLMs exhibit spontaneous cross-lingual alignment. However, although LLMs show promising cross-lingual alignment in IE, a significant imbalance across languages persists, highlighting an underlying deficiency. To address this, we propose KnowCoder-X, a powerful code LLM with advanced cross-lingual and multilingual capabilities for universal information extraction. Firstly, it standardizes the representation of multilingual schemas using Python classes, ensuring a consistent ontology across different languages. Then, IE across languages is formulated as a unified code generation task. Secondly, we enhance the model’s cross-lingual transferability through IE cross-lingual alignment instruction tuning on a translated instance prediction task we proposed. During this phase, we also construct a high-quality and diverse bilingual IE parallel dataset with 257k samples, called ParallelNER, synthesized by our proposed robust three-stage pipeline, with manual annotation to ensure quality. Although without training in 29 unseen languages, KnowCoder-X surpasses ChatGPT by 30.17% and SoTA by 20.03%, thereby demonstrating superior cross-lingual IE capabilities. Comprehensive evaluations on 64 IE benchmarks in Chinese and English under various settings demonstrate that KnowCoder-X significantly enhances cross-lingual IE transfer through boosting the IE alignment. Our code and dataset are available at: https://***/ICT-GoKnow/KnowCoder. © 2024, CC BY-NC-SA.

关键词： Python

Multimodal data fusion algorithm applied to robots 2

学校读者我要写书评

暂无评论

Multimodal data fusion algorithm applied to robots

2019 2nd International Conference on Computer Information Science and Artificial Intelligence, CISAI 2019

作者： Zhang, Xin Feng, Zhiquan Tian, Jinglan Yang, Xiaohui College of Information Science and Engineering University of Jinan Jinan Shandong250022 China Shandong Provincial Key Laboratory of Network Environment Intelligent Computing Technology University of Jinan Jinan Shandong250022 China

In recent years, the use of multimodal human-computer interaction technology to achieve the enhancement of human intelligence has become a new topic in human-computer interaction research. When the robot can't react correctly in a single mode, it is necessary to realize multimodal fusion. To this end, this paper proposes a multimodal fusion algorithm that applies the data obtained by the CNN feature layer to the decision-level. The speech recognition text is semantically matched with the text in the text library, and the similar probability vector is returned. At the same time, the similarity probability vector of the gesture recognition is obtained, and the data is filtered by the threshold, and the set of high probability data codes is assigned to the two modes. The intersection operation, and the final instruction is sent to the robot. The experimental results show that the influence of environmental factors on the single channel result is reduced, and the single mode ambiguity problem is eliminated. The multi-channel fusion algorithm with additional weight is more accurate than the common multi-channel fusion algorithm. At the same time, it has also been well received by many test users. © Published under licence by IOP Publishing Ltd.

关键词： Speech recognition

Cache Design Effect on Microarchitecture security: A Contrast between Xuantie-910 and BOOM

学校读者我要写书评

暂无评论

Cache Design Effect on Microarchitecture Security: A Contras...

IEEE International Conference on Trust, security and Privacy in computing and Communications (TrustCom)

作者： Zhe Zhou Xiaoyu Cheng Yang Sun Fang Jiang Fei Tong Yuxing Mao Ruilin Wang Southeast University Nanjing China Purple Mountain Laboratories Nanjing China School of Electrical Engineering Chongqing University Chongqin China State Key Laboratory of Power Transmission Equipment & System Security and New Technology Chongqing University Chongqing China School of Computing Newcastle University United Kingdom

Modern processors make use of optimization techniques such as cache and speculation mechanisms to greatly improve performance. But recent research has found that these techniques can also be exploited by attackers to perform powerful side-channel attacks. A large number of powerful cache-based attacks have been replicated and enhanced over Intel X86- and ARM-based architectures, but there is a relative lack of research on RISC-V-based architectures. Xuantie-910 and BOOM are both RISC-V-based processors. So far, cache-side channels in the unprivileged case of Xuantie-910 have not been proven, while cache attacks against BOOM are proliferating. There are two types of caches, including physically-indexed physically-tagged (PIPT) cache (adopted by Xuantie-910) and virtually-indexed physically-tagged (VIPT) cache (adopted by BOOM), corresponding to two different cache addressing forms. VIPT has higher addressing performance than PIPT, since it can directly obtain cache line index from virtual address. In this paper, we study Xuantie-910 and BOOM to explore the impact of cache design on the security of RISC-V-based microarchitecture. Specifically, we compare the impact of their cache addressing forms on precise flushing of cache lines at specified locations, which plays an important role in cache side-channel attacks. Experimental results show that for the VIPT cache in BOOM, the location-specified cache lines can be accurately flushed, and Spectre attack can be successfully carried out by using the cache side-channel. On the other hand, for the PIPT cache in Xuantie-910, it is impossible for attackers to directly and accurately flush the specified location of cache without affecting performance, which hinders the success of cache side-channel attacks. This provides us with an insight that one can adopt a VIPT-based cache with a mechanism similar to PIPT for preventing the accurate access of cache line index, which can not only keep the advantage of high-performance addre

关键词： Privacy Microarchitecture Program processors Side-channel attacks security Indexes Optimization

Deep learning-based forgery attack on document images

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhao, Lin Chen, Changsheng Huang, Jiwu the Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Media Security National Engineering Laboratory for Big Data System Computing Technology College of Electronics and Information Engineering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society China

With the ongoing popularization of online services, the digital document images have been used in various applications. Meanwhile, there have emerged some deep learning-based text editing algorithms which alter the textual information of an image in an end-to-end fashion. In this work, we present a low-cost document forgery algorithm by the existing deep learning-based technologies to edit practical document images. To achieve this goal, the limitations of existing text editing algorithms towards complicated characters and complex background are addressed by a set of network design strategies. First, the unnecessary confusion in the supervision data is avoided by disentangling the textual and background information in the source images. Second, to capture the structure of some complicated components, the text skeleton is provided as auxiliary information and the continuity in texture is considered explicitly in the loss function. Third, the forgery traces induced by the text editing operation are mitigated by some post-processing operations which consider the distortions from the print-and-scan channel. Quantitative comparisons of the proposed method and the exiting approach have shown the advantages of our design by reducing the about 2/3 reconstruction error measured in MSE, improving reconstruction quality measured in PSNR and in SSIM by 4 dB and 0.21, respectively. Qualitative experiments have confirmed that the reconstruction results of the proposed method are visually better than the existing approach in both complicated characters and complex texture. More importantly, we have demonstrated the performance of the proposed document forgery algorithm under a practical scenario where an attacker is able to alter the textual information in an identity document using only one sample in the target domain. The forged-and-recaptured samples created by the proposed text editing attack and recapturing operation have successfully fooled some existing document authenticat

关键词： Textures

CubeFlow: Money laundering detection with coupled tensors

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Sun, Xiaobing Zhang, Jiabao Zhao, Qiming Liu, Shenghua Chen, Jinglei Zhuang, Ruoyu Shen, Huawei Cheng, Xueqi CAS Key Laboratory of Network Data Science and Technology Insititute of Computing Technology Chinese Academy of Sciences China University of Chinese Academy of Sciences Beijing China China Construction Bank Fintech China

Money laundering (ML) is the behavior to conceal the source of money achieved by illegitimate activities, and always be a fast process involving frequent and chained transactions. How can we detect ML and fraudulent activity in large scale attributed transaction data (i.e. tensors)? Most existing methods detect dense blocks in a graph or a tensor, which do not consider the fact that money are frequently transferred through middle accounts. CubeFlow proposed in this paper is a scalable, flow-based approach to spot fraud from a mass of transactions by modeling them as two coupled tensors and applying a novel multi-attribute metric which can reveal the transfer chains accurately. Extensive experiments show CubeFlow outperforms state-of-the-art baselines in ML behavior detection in both synthetic and real data. Copyright © 2021, The Authors. All rights reserved.

关键词： Tensors

Real-Time multimodal emotion recognition system based on elderly accompanying robot 2

学校读者我要写书评

暂无评论

Real-Time multimodal emotion recognition system based on eld...

2019 2nd International Conference on Computer Information Science and Artificial Intelligence, CISAI 2019

作者： Dou, Shaosong Feng, Zhiquan Yang, Xiaohui Tian, Jinglan College of Information Science and Engineering University of Jinan Jinan Shandong250022 China Shandong Provincial Key Laboratory of Network Environment Intelligent Computing Technology University of Jinan Jinan Shandong250022 China

In recent years, people have been paying more attention to the impact of technological development on human emotion recognition. At the same time, China has become the country with the most elderly population in the world. However, due to the lack of real-Time multimodal emotion recognition technology for the elderly accompany robots, this paper proposes a deep learning decision-level fusion of real-Time emotion analysis model which is based on the background of the elderly care. The results of image and audio recognition are used for intersection and union operation to get emotional classification result, and the obtained emotional result is corresponding to the feedback behavior of the accompany *** after the experiment, the recognition algorithm proposed in this paper accuracy can reach about 90%, which is nearly 10% higher than the single mode, and the feedback of the robot has achieved the expected effect. © Published under licence by IOP Publishing Ltd.

关键词： Robots

A Review of Multimodal Explainable Artificial Intelligence: Past, Present and Future

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Sun, Shilin An, Wenbin Tian, Feng Nan, Fang Liu, Qidong Liu, Jun Shah, Nazaraf Chen, Ping School of Computer Science and Technology Xi’an Jiaotong University Xi’an710049 China Ministry of Education Key Laboratory of Intelligent Networks and Network Security Xi’an Jiaotong University Xi’an710049 China Faculty of Electronic and Information Engineering Xi’an Jiaotong University Xi’an710049 China Shaanxi Province Key Laboratory of Big Data Knowledge Engineering Xi’an Jiaotong University Xi’an710049 China Institute for Future Transport and Cities Coventry University Priory Street CoventryCV1 5FB United Kingdom Department of Engineering University of Massachusetts Boston BostonMA02125 United States

Artificial intelligence (AI) has rapidly developed through advancements in computational power and the growth of massive datasets. However, this progress has also heightened challenges in interpreting the "black-box" nature of AI models. To address these concerns, eXplainable AI (XAI) has emerged with a focus on transparency and interpretability to enhance human understanding and trust in AI decision-making processes. In the context of multimodal data fusion and complex reasoning scenarios, the proposal of Multimodal eXplainable AI (MXAI) integrates multiple modalities for prediction and explanation tasks. Meanwhile, the advent of Large Language Models (LLMs) has led to remarkable breakthroughs in natural language processing, yet their complexity has further exacerbated the issue of MXAI. To gain key insights into the development of MXAI methods and provide crucial guidance for building more transparent, fair, and trustworthy AI systems, we review the MXAI methods from a historical perspective and categorize them across four eras: traditional machine learning, deep learning, discriminative foundation models, and generative LLMs. We also review evaluation metrics and datasets used in MXAI research, concluding with a discussion of future challenges and directions. A project related to this review has been created at https://***/ShilinSun/mxai_review. Copyright © 2024, The Authors. All rights reserved.

关键词： Data fusion

Domain generalization for document authentication against practical recapturing attacks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Chen, Changsheng Zhang, Shuzheng Lan, Fengbo Huang, Jiwu The Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Media Security National Engineering Laboratory for Big Data System Computing Technology College of Electronics and Information Engineering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society China

Recapturing attack can be employed as a simple but effective anti-forensic tool for digital document images. Inspired by the document inspection process that compares a questioned document against a reference sample, we proposed a document recapture detection scheme by employing Siamese network to compare and extract distinct features in a recapture document image. The proposed algorithm takes advantages of both metric learning and image forensic techniques. Instead of adopting Euclidean distance-based loss function, we integrate the forensic similarity function with a triplet loss and a normalized softmax loss. After training with the proposed triplet selection strategy, the resulting feature embedding clusters the genuine samples near the reference while pushes the recaptured samples apart. In the experiment, we consider practical domain generalization problems, such as the variations in printing/imaging devices, substrates, recapturing channels, and document types. To evaluate the robustness of different approaches, we benchmark some popular off-the-shelf machine learning-based approaches, a state-of-the-art document image detection scheme, and the proposed schemes with different network backbones under various experimental protocols. Experimental results show that the proposed schemes with different network backbones have consistently outperformed the state-of-the-art approaches under different experimental settings. Specifically, under the most challenging scenario in our experiment, i.e., evaluation across different types of documents that produced by different devices, we have achieved less than 5.00% APCER (Attack Presentation Classification Error Rate) and 5.56% BPCER (Bona Fide Presentation Classification Error Rate) by the proposed network with ResNeXt101 backbone at 5% BPCER decision *** Codes 94A08 Copyright © 2021, The Authors. All rights reserved.

关键词： Feature extraction

A Fault Detection Method for Tree Contact Single-Phase-To-Ground Fault

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Wu, Juping Xv, Huikai Jin, Min Gong, Guoliang Chen, Gang Mao, Wenyu Lu, Huaxiang Chen, Tianxiang Institute of Semiconductors CAS Beijing100083 China University of Chinese Academy of Sciences Beijing China Semiconductor Neural Network Intelligent Perception and Computing Technology Beijing Key Laboratory Beijing China Collage of Microelectronics University of Chinese Academy of Sciences Beijing China Materials and Optoelectronics Research Center University of Chinese Academy of Sciences Beijing China Tianjin University of Science and Technology Tianjin300457 China College of Nuclear Technology and Automation Engineering Chengdu University of Technology Chengdu610059 China

High impedance fault(HIF), especially tree contact single-phase-to-ground fault(TSF), often occurs in resonant grounding systems and are difficult to detect. Therefore, there is an urgent need for an effective TSF detection technology. In this paper, by analyzing the equivalent circuits of different impedance faults in resonant grounding system, and analyzing the transient zero-sequence components of underdamped state and overdamped state under different impedance, the conclusion is drawn: When a TSF occurs, the transient zero-sequence voltage will have a unique " Beat frequency " characteristic, and the problem of covering this phenomenon due to the imbalance of distribution network is effectively solved by fast independent component analysis(FastICA). Based on this conclusion, A fault detection method for TSF based on FastICA is proposed. Finally, the study conducted real experiments using two different impedance trees, bamboo and pine, to verify the proposed detection method. The results of the experiment found clear support for that the detection method is accurate and effective. © 2023, The Authors. All rights reserved.

关键词： Fault detection