检索结果-内蒙古大学图书馆

Evolutionary Computation in Action: Hyperdimensional Deep Embedding Spaces of Gigapixel Pathology Images

ieee TRANSACTIONS ON EVOLUTIONARY COMPUTATION 2023年第1期27卷 52-66页

作者： Bidgoli, Azam Asilian Rahnamayan, Shahryar Dehkharghanian, Taher Riasatian, Abtin Tizhoosh, Hamid R. Ontario Tech Univ Nat Inspired Computat Intelligence Lab Oshawa ON L1G 0C5 Canada McMaster Univ Dept Pathol & Mol Med Hamilton ON L8S 4L8 Canada Univ Waterloo Lab Knowledge Inference Med Image Anal Waterloo ON N2L 3G1 Canada Mayo Clin Dept Artificial Intelligence & Informat Rochester MN 55905 USA

One of the main obstacles of adopting digital pathology is the challenge of efficient processing of hyperdimensional digitized biopsy samples, called whole slide images (WSIs). Exploiting deep learning and introducing compact WSI representations are urgently needed to accelerate image analysis and facilitate the visualization and interpretability of pathology results in a postpandemic world. In this article, we introduce a new evolutionary approach for WSI representation based on large-scale multiobjective optimization (LSMOP) of deep embeddings. We start with patch-based sampling to feed KimiaNet, a histopathology-specialized deep network, and to extract a multitude of feature vectors. Coarse multiobjective feature selection uses the reduced search space strategy guided by the classification accuracy and the number of features. In the second stage, the frequent features histogram (FFH), a novel WSI representation, is constructed by multiple runs of coarse LSMOP. Fine evolutionary feature selection is then applied to find a compact (short-length) feature vector based on the FFH and contributes to a more robust deep-learning approach to digital pathology supported by the stochastic power of evolutionary algorithms. We validate the proposed schemes using The Cancer Genome Atlas (TCGA) images in terms of WSI representation, classification accuracy, and feature quality. Furthermore, a novel decision space for multicriteria decision making in the LSMOP field is introduced. Finally, a patch-level visualization approach is proposed to increase the interpretability of deep features. The proposed evolutionary algorithm finds a very compact feature vector to represent a WSI (almost 14 000 times smaller than the original feature vectors) with 8% higher accuracy compared to the codes provided by the state-of-the-art methods

关键词： Feature extraction Optimization Evolutionary computation data mining Histopathology Cancer Task analysis Deep neural network (DNN) digital pathology dimension reduction evolutionary computation (EC) feature selection image representation innovization large-scale multiobjective optimization (LSMOP) transfer learning

来源：评论

学校读者我要写书评

暂无评论

Modeling and Validating System Dynamics in Saudi Synthetic Electric Grid 56

Modeling and Validating System Dynamics in Saudi Synthetic E...

引用

56th North American Power symposium

作者： Alonize, Mishal Wang, Jinning Malkmus, Zackery Li, Fangxing (Fran) Univ Tennessee Dept Elect Engn & Comp Sci Knoxville TN 37996 USA

ISBN: (纸本)9798331521042;9798331521035

This paper introduces a modeling approach for creating a synthetic Saudi power grid, designed to simulate the transient dynamics of a real power grid. This synthetic grid model, created from open-source data, is available for open sharing and can be utilized for educational, and research purposes. The cases are built to match a detailed model of Saudi power system elements, Saudi code requirements, and statistical characteristics found in actual power grids. First, we describe how we built, tested, tuned, and validated this synthetic Saudi grid model by using statistics from currently available data. The CURENT large-scale Testbed (LTB) and its ANDES package enabled power flow calculations, simulations of transient stability, and analysis of small-signal stability in transmission systems. Second, we describe our use of the ANDES package for dynamic simulation and testing, and our use of the LTB's AGVis package for grid visualization. This Saudi synthetic grid model has 223 buses that have been tested, tuned and validated to ensure that they are realistic and are practical for various studies. Also, it greatly enhances the diversity of existing collections of synthetic grids in the current literature.

关键词： Synthetic grid grid code requirements generator dynamics model and validation large-scale Testbed

来源：评论

学校读者我要写书评

暂无评论

FedSH: Towards Privacy-Preserving Text-Based Person Re-Identification

引用

ieee TRANSACTIONS ON MULTIMEDIA 2024年 26卷 5065-5077页

作者： Ma, Wentao Wu, Xinyi Zhao, Shan Zhou, Tongqing Guo, Dan Gu, Lichuan Cai, Zhiping Wang, Meng Anhui Agr Univ Sch Informat & Artificial Intelligence Hefei 230036 Peoples R China Natl Univ Def Technol Coll Comp Changsha 410073 Peoples R China Hefei Univ Technol Sch Comp & Informat Engn Hefei 230009 Peoples R China

Text-based person re-identification (ReID) has enabled canonical applications in searching for and tracking targets from large-scale surveillance images with textual descriptions. Yet, existing text-based person ReID systems employ centralized model training that gathers images captured by different institutes' cameras into one place, which poses severe privacy threats to sensitive institutional information. This work is then devoted to exploring privacy-preserving text-based person ReID and proposes the framework of FedSH by tailoring the federated learning paradigm for distributed searching knowledge extraction. Specifically, FedSH resolves the local model generalization and entity boundary obscuring limitations, caused by inner-institute data homogeneity and inter-institute data heterogeneity, via building multi-granularity feature representation and a semantically self-aligned network. Meanwhile, it reduces the communication burden introduced by the embedding for multiple modals by updating common representation subspaces during federated learning. Experimental results on two public benchmarks demonstrate that our method can achieve at most 16.47% and 16.02% person ReID performance improvement by the Rank-1 metric, compared with 6 State-of-The-Art (SoTA) baselines and 6 ablation studies. We believe that our work will inspire the community to investigate the potential of implementing Federated Learning in real-world image retrieval and ReID scenarios.

关键词： Semantics Training Task analysis Privacy visualization Federated learning Servers Text-based Person ReID Cross-modal Retrieval Federated Learning Multi-granularity Representation

来源：评论

学校读者我要写书评

暂无评论

Visual Parameter Selection for Spatial Blind Source Separation

引用

COMPUTER GRAPHICS FORUM 2022年第3期41卷 157-168页

作者： Piccolotto, N. Boegl, M. Muehlmann, C. Nordhausen, K. Filzmoser, P. Miksch, S. TU Wien Inst Visual Comp & Human Ctr Technol Vienna Austria TU Wien Inst Stat & Math Methods Econ Vienna Austria Univ Jyvaskyla Jyvaskyla Finland

analysis of spatial multivariate data, i.e., measurements at irregularly-spaced locations, is a challenging topic in visualization and statistics alike. Such data are inteGral to many domains, e.g., indicators of valuable minerals are measured for mine prospecting. Popular analysis methods, like PCA, often by design do not account for the spatial nature of the data. Thus they, together with their spatial variants, must be employed very carefully. Clearly, it is preferable to use methods that were specifically designed for such data, like spatial blind source separation (SBSS). However, SBSS requires two tuning parameters, which are themselves complex spatial objects. Setting these parameters involves navigating two large and interdependent parameter spaces, while also taking into account prior knowledge of the physical reality represented by the data. To support analysts in this process, we developed a visual analytics prototype. We evaluated it with experts in visualization, SBSS, and geochemistry. Our evaluations show that our interactive prototype allows to define complex and realistic parameter settings efficiently, which was so far impractical. Settings identified by a non-expert led to remarkable and surprising insights for a domain expert. Therefore, this paper presents important first steps to enable the use of a promising analysis method for spatial multivariate data.

关键词： CCS Concepts • Human-centered computing → visualization techniques GeoGraphic visualization

来源：评论

学校读者我要写书评

暂无评论

Amortised Encoding for large High-Resolution Displays 11

Amortised Encoding for Large High-Resolution Displays

引用

ieee 11th symposium on large data analysis and visualization (LDAV)

作者： Friess, Florian Braun, Matthias Reina, Guido Ertl, Thomas Univ Stuttgart Stuttgart Germany

ISBN: (纸本)9781665432832

Both visual detail and a low-latency transfer of image data are required for collaborative exploration of scientific data sets across large high-resolution displays. In this work, we present an approach that reduces the resolution before the encoding and uses temporal upscaling to reconstruct the full resolution image, reducing the overall latency and the required bandwidth without significantly impacting the details perceived by observers. Our approach takes advantage of the fact that humans do not perceive the full details of moving objects by providing a perfect reconstruction for static parts of the image, while non-static parts are reconstructed with a lower quality. This strategy enables a substantial reduction of the encoding latency and the required bandwidth with barely noticeable changes in visual quality, which is crucial for collaborative analysis across display walls at different locations. Additionally, our approach can be combined with other techniques aiming to reduce the required bandwidth while keeping the quality as high as possible, such as foveated encoding. We demonstrate the reduced overall latency, the required bandwidth, as well as the high image quality using different visualisations.

关键词： Human-centered computing visualization systems and tools Computing methodologies Graphics systems and interfaces

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Font Generation With Weakly Supervised Localized Representations

引用

ieee TRANSACTIONS ON PATTERN analysis AND MACHINE INTELLIGENCE 2024年第3期46卷 1479-1495页

作者： Park, Song Chun, Sanghyuk Cha, Junbum Lee, Bado Shim, Hyunjung NAVER AI Lab Seongnam 13561 South Korea NAVER CLOVA Seongnam 13561 South Korea Kakao Brain Seongnam 63309 South Korea Korea Adv Inst Sci & Technol Kim Jaechul Grad Sch AI Daejeon 34141 South Korea

Automatic few-shot font generation aims to solve a well-defined, real-world problem because manual font designs are expensive and sensitive to the expertise of designers. Existing methods learn to disentangle style and content elements by developing a universal style representation for each font style. However, this approach limits the model in representing diverse local styles because it is unsuitable for complicated letter systems. For example, Chinese characters consist of a varying number of components (often called "radical") with a highly complex structure. In this paper, we propose a novel font generation method that learns localized styles, namely component-wise style representations, instead of universal styles. The proposed style representations enable synthesizing complex local details in text designs. However, learning component-wise styles solely from a few reference glyphs is infeasible when a target script has a large number of components, for example, over 200 for Chinese. To reduce the number of required reference glyphs, we represent component-wise styles by a product of component and style factors inspired by low-rank matrix factorization. Owing to the combination of strong representation and a compact factorization strategy, our method shows remarkably better few-shot font generation results (with only eight reference glyphs) than other state-of-the-art methods. Moreover, strong locality supervision was not utilized, such as the location of each component, skeleton, or strokes.

关键词： Task analysis Libraries Feature extraction data models visualization Training Skeleton Few-shot font generation font generation few-shot generation image-to-image translation computer vision

来源：评论

学校读者我要写书评

暂无评论

From Chatbots to Phishbots?: Phishing Scam Generation in Commercial large Language Models 45

From Chatbots to Phishbots?: Phishing Scam Generation in Com...

引用

45th ieee symposium on Security and Privacy (SP)

作者： Roy, Sayak Saha Thota, Poojitha Naragam, Krishna Vamsi Nilizadeh, Shirin Univ Texas Arlington Arlington TX 76019 USA

ISBN: (纸本)9798350331318;9798350331301

The advanced capabilities of large Language Models (LLMs) have made them invaluable across various applications, from conversational agents and content creation to data analysis, research, and innovation. However, their effectiveness and accessibility also render them susceptible to abuse for generating malicious content, including phishing attacks. This study explores the potential of using four popular commercially available LLMs, i.e., ChatGPT (GPT 3.5 Turbo), GPT 4, Claude, and Bard, to generate functional phishing attacks using a series of malicious prompts. We discover that these LLMs can generate both phishing websites and emails that can convincingly imitate well-known brands and also deploy a range of evasive tactics that are used to elude detection mechanisms employed by anti-phishing systems. These attacks can be generated using unmodified or "vanilla" versions of these LLMs without requiring any prior adversarial exploits such as jailbreaking. We evaluate the performance of the LLMs towards generating these attacks and find that they can also be utilized to create malicious prompts that, in turn, can be fed back to the model to generate phishing scams - thus massively reducing the prompt-engineering effort required by attackers to scale these threats. As a countermeasure, we build a BERT-based automated detection tool that can be used for the early detection of malicious prompts to prevent LLMs from generating phishing content. Our model is transferable across all four commercial LLMs, attaining an average accuracy of 96% for phishing website prompts and 94% for phishing email prompts. We also disclose the vulnerabilities to the concerned LLMs, with Google acknowledging it as a severe issue. Our detection model is available for use at Hugging Face, as well as a ChatGPT Actions plugin.

关键词： ChatGPT Detection tool large Language Models Phishing Prompt Engineering vulnerabilities

来源：评论

学校读者我要写书评

暂无评论

Learning Objectives, Insights, and Assessments: How Specification Formats Impact Design

引用

ieee TRANSACTIONS ON visualization AND COMPUTER GRAPHICS 2022年第1期28卷 676-685页

作者： Lee-Robbins, Elsie He, Shiqing Adar, Eytan Univ Michigan Sch Informat Ann Arbor MI 48109 USA

Despite the ubiquity of communicative visualizations, specifying communicative intent during design is ad hoc. Whether we are selecting from a set of visualizations, commissioning someone to produce them, or creating them ourselves, an effective way of specifying intent can help guide this process. Ideally, we would have a concise and shared specification language. In previous work, we have argued that communicative intents can be viewed as a learning/assessment problem (i.e., what should the reader learn and what test should they do well on). Learning-based specification formats are linked (e.g., assessments are derived from objectives) but some may more effectively specify communicative intent. Through a large-scale experiment, we studied three specification types: learning objectives, insights, and assessments. Participants, guided by one of these specifications, rated their preferences for a set of visualization designs. Then, we evaluated the set of visualization designs to assess which specification led participants to prefer the most effective visualizations. We find that while all specification types have benefits over no-specification, each format has its own advantages. Our results show that learning objective-based specifications helped participants the most in visualization selection. We also identify situations in which specifications may be insufficient and assessments are vital.

关键词： visualization Task analysis data visualization Stakeholders Usability Taxonomy Interviews Communicative visualization evaluation visualization specification

来源：评论

学校读者我要写书评

暂无评论

UNITED STATES WEST COAST SURFACE DEFORMATION WITH WIDE-SWATH L-BAND ALOS-2 PALSAR-2 AND NISAR

UNITED STATES WEST COAST SURFACE DEFORMATION WITH WIDE-SWATH...

引用

ieee International Geoscience and Remote Sensing symposium (IGARSS)

作者： Aissa, S. Belhadj Fielding, E. J. Liu, Z. Tymofyeyeva, E. Havazli, E. Rosen, P. Simons, Mark Lindsay, D. Burgmann, R. Bawden, G. W. CALTECH Jet Prop Lab Pasadena CA 91125 USA CALTECH Seismol Lab Pasadena CA USA Univ Calif Berkeley Dept Earth & Planetary Sci Berkeley CA USA NASA HQ Washington DC USA

ISBN: (纸本)9798350360332;9798350360325

Launching in early 2024, the NASA-ISRO SAR (NISAR) mission will provide global data freely accessible enabling large scale surface deformation monitoring with synthetic aperture radar (SAR) acquired at L-band and S-band radar wavelengths. In preparation for calibration and validation of the NISAR L-band data, the NISAR Solid earth science team is systematically processing over 450 Japan Aerospace Exploration Agency (JAXA) ALOS-2 PalSAR-2 wide-swath (ScanSAR) L-band acquisitions covering the West Coast of the United States for measuring co-seismic, secular and transient displacements. The area spans California, Washington and Oregon.

关键词： NISAR ALOS-2 PALSAR Land motion SAR interferometry InSAR time series analysis

来源：评论

学校读者我要写书评

暂无评论

data Augmentation based on Inverse Transform Sampling for Improved Tissue Classification via Electrical Impedance Spectroscopy 18

Data Augmentation based on Inverse Transform Sampling for Im...

引用

18th ieee Sensors Applications symposium (SAS)

作者： McDermott, Conor Rossa, Carlos Carleton Univ Syst & Comp Engn Ottawa ON Canada

ISBN: (纸本)9798350323078

Ultrasound-guided needle biopsy often leads to uncertain results due to poor visualization of targets in the image. Biopsy needles equipped with electrical impedance spectroscopy (EIS) sensors can be used to identify the tissue at the needle tip and provide more accurate needle guidance and lesion targeting. A machine learning algorithm is often used to classify the tissue based on the measured EIS spectrum. However, training machine learning algorithms requires large amounts of data, which is rarely available in biomedical applications. A solution to increase the size of the dataset and improve the performance of the classifier is to create synthetic data that closely mimics the original measured data. This paper proposes inverse transform sampling as a data augmentation method for EIS to bolster training dataset size. It exploits the cumulative distribution functions of the target data and uses inverse sampling to generate new, synthetic data. The method is demonstrated using an EIS dataset composing 13 different ex-vivo tissue types. The method is then validated by comparing the performance of the synthetic data to the original data through the use of an artificial neural network (ANN) and a convolutional neural network (CNN). The classification results indicate that classification sensitivity, precision, and accuracy increase by at least 27.38, 24.86, and 19.41%, respectively, when the classifiers used a mix of original data and data augmented with the proposed method.

关键词： Inverse problems

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：