检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Zhao, Weilin Huang, Yuxiang Han, Xu Liu, Zhiyuan Zhang, Zhengyan Li, Kuai Chen, Chen Yang, Tao Sun, Maosong Department of Computer Science and Technology Institute for Artificial Intelligence Beijing National Research Center for Information Science and Technology Tsinghua University Beijing China Quan Cheng Laboratory Shandong China Tencent Machine Learning Platform Department Tencent Inc Beijing China

Recently, there has been a demand to deploy Large Language Models (LLMs) on personal devices such as laptops and smartphones. These LLMs have different model variants when handling different tasks. However, personal devices have limited resources and require reduced storage overhead. To address this, there are two key methods available: the first is model compression, which compresses LLMs into smaller sizes;the second is LoRA, which can transfer an LLM to other tasks with very few parameters, avoiding the storage of multiple model variants in multi-task scenarios by only preserving LoRAs. However, our experiments show that directly combining these two methods yields sub-optimal performance. Considering that the open-source community has already contributed many LoRAs to LLMs, we propose to adapt these existing LoRAs from the LLMs to their compressed version and introduce a Compression-Aware LoRA (CA-LoRA) framework. We incorporate knowledge inheritance and recovery strategies to recover the lost knowledge caused by model compression. Experiment results demonstrate that CA-LoRA outperforms the vanilla LoRA methods applied to a compressed LLM and achieves comparable performance to the non-compressed LLM with existing LoRA modules. The source code of CA-LoRA is available at https://***/thunlp/CA-LoRA. Copyright © 2023, The Authors. All rights reserved.

关键词： Open systems

来源：评论

学校读者我要写书评

暂无评论

Towards Adversarial Robustness for Multi-Mode Data through Metric Learning

引用

Sensors 2023年第13期23卷 6173-6173页

作者： Khan, Sarwar Chen, Jun-Cheng Liao, Wen-Hung Chen, Chu-Song Research Center for Information Technology Innovation Academia Sinica Taipei11529 Taiwan Social Networks Human-Centered Computing Taiwan International Graduate Program Academia Sinica Taipei11529 Taiwan Department of Computer Science National Chengchi University Taipei11605 Taiwan Department of Computer Science and Information Engineering National Taiwan University Taipei106319 Taiwan

Adversarial attacks have become one of the most serious security issues in widely used deep neural networks. Even though real-world datasets usually have large intra-variations or multiple modes, most adversarial defense methods, such as adversarial training, which is currently one of the most effective defense methods, mainly focus on the single-mode setting and thus fail to capture the full data representation to defend against adversarial attacks. To confront this challenge, we propose a novel multi-prototype metric learning regularization for adversarial training which can effectively enhance the defense capability of adversarial training by preventing the latent representation of the adversarial example changing a lot from its clean one. With extensive experiments on CIFAR10, CIFAR100, MNIST, and Tiny ImageNet, the evaluation results show the proposed method improves the performance of different state-of-the-art adversarial training methods without additional computational cost. Furthermore, besides Tiny ImageNet, in the multi-prototype CIFAR10 and CIFAR100 where we reorganize the whole datasets of CIFAR10 and CIFAR100 into two and ten classes, respectively, the proposed method outperforms the state-of-the-art approach by 2.22% and 1.65%, respectively. Furthermore, the proposed multi-prototype method also outperforms its single-prototype version and other commonly used deep metric learning approaches as regularization for adversarial training and thus further demonstrates its effectiveness. © 2023 by the authors.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Intriguing Effect of the Correlation Prior on ICD-9 Code Assignment 61

Intriguing Effect of the Correlation Prior on ICD-9 Code Ass...

引用

61st Annual Meeting of the Association for Computational Linguistics, ACL 2023

作者： Yang, Zihao Zhang, Chenkang Wu, Muru Liu, Xujin Chris Jiang, Lavender Yao Cho, Kyunghyun Oermann, Eric Karl Center for Data Science New York University United States Department of Neurosurgery NYU Langone Health United States Department of Electrical and Computer Engineering NYU Tandon School of Engineering United States Courant Institute of Mathematical Sciences New York University United States Canadian Institute for Advanced Research Canada Prescient Design United States Department of Radiology NYU Langone Health United States Neuroscience Institute NYU Langone Health United States

ISBN: (纸本)9781959429692

The Ninth Revision of the International Classification of Diseases (ICD-9) is a standardized coding system used to classify health conditions. It is used for billing, tracking individual patient conditions, and for epidemiology. The highly detailed and technical nature of the codes and their associated medical conditions make it difficult for humans to accurately record them. researchers have explored the use of neural networks, particularly language models, for automated ICD-9 code assignment. However, the imbalanced distribution of ICD-9 codes leads to poor performance. One solution is to use domain knowledge to incorporate a useful prior. This paper evaluates the usefulness of the correlation bias: we hypothesize that correlations between ICD-9 codes and other medical codes could help improve language models' performance. We showed that while the correlation bias worsens the overall performance, the effect on individual class can be negative or positive. Performance on classes that are more imbalanced and less correlated with other codes is more sensitive to incorporating the correlation bias. This suggests that while the correlation bias has potential to improve ICD-9 code assignment in certain cases, the applicability criteria need to be more carefully studied. © 2023 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Computational Methods for Analysing Biometric Systems 3

Computational Methods for Analysing Biometric Systems

引用

3rd International Conference on Technological Advancements in Computational sciences, ICTACS 2023

作者： Multani, Fahim Bijlwan, Sheela Upadhyay, Viyat Varun Goyal, Bhawna Dogra, Ayush Lepcha, Dawa Chyophel Ies College of Technology Bhopal Department of Computer Science Engineering Madhya Pradesh India Uttaranchal School of Computing Sciences Uttaranchal Institute of Technology Uttaranchal University Dehradun248007 India Gla University Department of Mechanical Engineering Mathura Mathura India University Center for Research and Development Chandigarh University Punjab Mohali India Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Chandigarh University Department Electronics and Communication Engineering Punjab Mohali India

ISBN: (纸本)9798350342338

The use of biometric data, such as various anatomical and behavioural attributes in automatic identification and authentication systems for access control, surveillance, computer security, and law enforcement are increasing. Several nations are already or will soon begin to use biometric technology. In this paper, there is a in detail analysis of innovative biometric systems and also the research study of how the computational techniques used intelligent data system in order to analyse biometric system, The efficacy of biometric technologies will depend on how well they work. Biometric system evaluation examines how well biometric systems adapt to particular characteristics. Planning and carrying out biometric tests are frequently handled by an impartial third party. A measure of confidence in the validity of the claim might be one of the options the algorithm offers. Applications that validate a user's address or confirm their identity at the point of sale are examples of verification applications. © 2023 IEEE.

关键词： Attributes Biometric data Computational techniques Innovative research Technology

来源：评论

学校读者我要写书评

暂无评论

Innovative Stacked Yellow and Blue Mini-LED Chip for White Lamp Applications

引用

Micromachines 2024年第6期15卷 796-796页

作者： Lee, Tzu-Yi Huang, Chien-Chi Miao, Wen-Chien Hsiao, Fu-He Tsai, Chia-Hung Hung, Yu-Ying Chen, Fang-Chung Lin, Chun-Liang Ohkawa, Kazuhiro He, Jr-Hau Hong, Yu-Heng Kuo, Hao-Chung Department of Photonics College of Electrical and Computer Engineering National Yang Ming Chiao Tung University Hsinchu30010 Taiwan Semiconductor Research Center Foxconn Research Taipei11492 Taiwan Department of Electrophysics College of Science National Yang Ming Chiao Tung University Hsinchu30010 Taiwan Thuwal23955-6900 Saudi Arabia Department of Materials Science and Engineering City University of Hong Kong Tat Chee Avenue Kowloon Hong Kong

This study introduces a novel approach for fabricating vertically stacked mini-LED arrays, integrating InGaN yellow and blue epitaxial layers with a stress buffer layer to enhance optoelectronic characteristics and structural stability. This method significantly simplifies the LED design by reducing the need for RGB configurations, thus lowering costs and system complexity. Employing vertical stacking integration technology, the design achieves high-density, efficient white light production suitable for multifunctional applications, including automotive lighting and outdoor signage. Experimental results demonstrate the exceptional performance of the stacked yellow and blue mini-LEDs in terms of luminous efficiency, wavelength precision, and thermal stability. The study also explores the performance of these LEDs under varying temperature conditions and their long-term reliability, indicating that InGaN-based yellow LEDs offer superior performance over traditional AlGaInP yellow LEDs, particularly in high-temperature environments. This technology promises significant advancements in the design and application of lighting systems, with potential implications for both automotive and general illumination markets. © 2024 by the authors.

关键词： Light emitting diodes

来源：评论

学校读者我要写书评

暂无评论

Enterprise Support Hierarchical Model with Secure Data Protocol 3

Enterprise Support Hierarchical Model with Secure Data Proto...

引用

3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023

作者： Vinay Simha Reddy, T. Parvathi, S. Sarada, Ch. Mohmmed, Karrar Hatif Sampath Dakshina Murthy, A. Saikumar, K. Malla Reddy University Dept of Artificial Intelligence and Machine Learning Maisammaguda Hyderabad Telangana Dulapally India Mlr Institute of Technology Department of Computer Science and Information Technology Telangana Hyderabad India Cvr College of Engineering Department of Cse Telangana Hyderabad India Scientific Research Center Al-Ayen University Thi-Qar64001 Iraq Department of Ece Andhra Pradesh Duvvada India Koneru Lakshmaiah Education Foundation Dept of Cse Telangana Hyderabad500075 India

ISBN: (纸本)9798350319125

CHARON is a cloud-upheld capacity framework that can store and share large information in a manner that is protected, solid, and proficient. It does this by utilizing different cloud sources and capacity areas to meet the lawful prerequisites for private individual information. CHARON has three things that make it stick out: It doesn't require trust in any one individual or thing. It needn't bother with a client-oversaw server. It handles large records well across a bunch of putting-away administrations in various regions of the planet. We likewise made another Byzantine-strong information-driven rental plan to stay away from composing issues between clients utilizing shared documents. We test CHARON by displaying test processes from science, a notable huge information region, utilizing miniature and application-based measures. The outcomes show that our novel plan isn't just imaginable, however, it additionally stirs up to multiple times better compared to other cloud-based arrangements from beginning to end. © 2023 IEEE.

关键词： CHARON Cloud computing

来源：评论

学校读者我要写书评

暂无评论

Deep Bispectral Analysis of Conversational Speech Towards Emotional Climate Recognition 5

Deep Bispectral Analysis of Conversational Speech Towards Em...

引用

5th IEEE International Conference on Artificial Intelligence in Engineering and Technology, IICAIET 2023

作者： Alhussein, Ghada Alkhodari, Mohanad Khandoker, Ahsan H. Hadjileontiadis, Leontios J. Khalifa University of Science and Technology Department of Biomedical Engineering P.O. Box 127788 Abu Dhabi United Arab Emirates Khalifa University of Science and Technology Healthcare Engineering Innovation Center P.O. Box 127788 Abu Dhabi United Arab Emirates University of Oxford Cardiovascular Clinical Research Facility Radcliffe Department of Medicine OxfordOX1 2JD United Kingdom Aristotle University of Thessaloniki Department of Electrical & Computer Engineering ThessalonikiGR 54124 Greece

ISBN: (纸本)9798350304152

Peers' conversational speech plays a significant role in shaping the emotional climate (EC) during interactions. Machine-based recognition of EC provides insights into the emotional perception of conversations by both peers and external observers. In this paper, we propose DeepBispec, a novel approach for EC recognition using deep bispectral analysis. DeepBispec applies windowed bispectral analysis to the 1D conversational speech signal. By capturing higher-order spectral correlations, the bispectrum magnifies the nonlinear characteristics present in speech signals. The estimated 2D -bispectrum magnitude contours, representing these interactions, are transformed into colored images and fed into a convolutional neural network (CNN). The CNN learns deep features from the bispectrum magnitude contours, enabling it to predict the valence (V) and arousal (A) labels associated with the EC. Evaluating DeepBispec on the K- EmoCon dataset using 10-fold cross-validation, we achieve an accuracy of 0.789 (A)/0.771 (V), an F1 score of 0.850 (A)/0.836 (V), and an area under the curve (AUC) of 0.812 (A)/0.788 (V). These results surpass existing benchmarks, demonstrating the effectiveness of bispectrum in capturing nonlinear characteristics and improving EC recognition. DeepBispec introduces an innovative approach to analyzing conversational speech for enhanced EC recognition. By leveraging deep bispectral analysis and CNN, it uncovers the higher-order spectral correlations and nonlinear dynamics of speech signals. This contributes to a deeper understanding of emotional dynamics in conversations and provides valuable insights into EC perception. © 2023 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Covert Communication in the D2D-Enabled Cellular Network with Multiple Non-Colluding Wardens

Covert Communication in the D2D-Enabled Cellular Network wit...

引用

International Conference on Networking and Network Applications (NaNA)

作者： Jingsen Jiao Huihui Wu Ranran Sun Qifeng Miao Yizhi Cao Hangzhou Institute of Technology Xidian University Hangzhou China School of Computer Science and Technology Xidian University Xi'an China Department of Automation Beijing National Research Center for Information Science and Technology (BNRist) Tsinghua University Beijing China

This work investigates the covert communication in a D2D (device-to-device)-enabled cellular network, where a cellular user CT wants to transmit a message covertly to another cellular user CR through a base station BS without being detected by multiple non-colluding wardens with the help of an underlaying D2D pair. Although D2D communication can confuse wardens, it can also interfere with the covert communication of the cellular link. To fully understand the impact of D2D communication on the covert communication of the cellular link, we first explore the average minimum detection error probability of non-colluding wardens. Then, the average covert rate of the cellular link is studied. Eventually, expansive numerical and simulation results are presented to demonstrate the correctness of our theoretical analysis and also to illustrate the impact of D2D communication on the covert communication of the cellular link.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Improvement of Jakarta's air quality during large scale social restriction 2

Improvement of Jakarta's air quality during large scale soci...

引用

2nd International Conference on Biospheric Harmony Advanced research, ICOBAR 2020

作者： Rahutomo, R. Purwandari, K. Sigalingging, J.W.C. Pardamean, B. Information System Department School of Information Systems Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program Master of Computer Science Program Bina Nusantara University Jakarta11480 Indonesia Database Center Division of BMKG Meteorological Climatological and Geophysical Agency Jakarta10720 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia

After Large-Scale Social Restriction (PSBB) established in Jakarta, a change of air quality was indicated by the citizens. Representatives of Indonesia's Agency for Meteorology, Climatology, and Geophysics (BMKG) supported as the fact that mobility of Jakarta's citizen and commuters were reduced. With utilizing Weather Data Acquisition System managed by BMKG, this research compared fluctuations of Jakarta's air quality during two periods of time: Pre-PSBB and During-PSBB. This study proved that Jakarta's air quality has positively improved compared to the previous year, although in the lower range of unhealthy level. © Published under licence by IOP Publishing Ltd.

关键词： Air quality

来源：评论

学校读者我要写书评

暂无评论

Dual-View Multimodal Interaction in Multimodal Sentiment Analysis

Dual-View Multimodal Interaction in Multimodal Sentiment Ana...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Kang Zhu Cunhang Fan Jianhua Tao Jun Xue Heng Xie Xuefei Liu Yongwei Li Zhengqi Wen Zhao Lv Anhui Province Key Laboratory of Multimodal Cognitive Computation School of Computer Science and Technology Anhui University Hefei China Department of Automation Tsinghua University China Beijing National Research Center for Information Science and Technology Tsinghua University China Chinese Academy of Sciences Institute of Automation

ISBN: (数字)9798350390155

ISBN: (纸本)9798350390162

Outstanding performance in sentiment analysis not only relies on the design of sophisticated fusion methods but also on the crucial step of designing excellent modal interaction methods. To the best of our knowledge, there are few methods addressing the capture of multimodal spatial features. Majority of feature interactions have been primarily focused on temporal aspects, with less attention given to the combined spatiotemporal feature interaction (SFI). In this paper, we design a dual-view multimodal interaction method, named DVMI, primarily consisting of two parts. In the first part, a triangular convolutional module is proposed for ample temporal interaction between modalities, implicit local and global SFI, and capturing global spatial representations. Building upon the foundation laid in the first part, the second part employs an attention mechanism for explicit global SFI. To demonstrate the effectiveness of the DVMI framework,we conduct extensive experiments on three datasets, achieving state-of-the-art experimental results.

关键词： Sentiment analysis Technological innovation Attention mechanisms Buildings Spatiotemporal phenomena

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：