检索结果-内蒙古大学图书馆

Adaptive pelican optimization with optimized mask RCNN for automatic lung cancer detection

Multimedia Tools and Applications 2025年 1-21页

作者： Sudha, R. Maheswari, K.M. Uma Department of Computer Science and Engineering SRM Institute of Science and Technology Chengalpattu India Department of Computing Technologies SRM Institute of Science and Technology Chengalpattu India

Lung cancer is a prevalent and deadly disease worldwide, necessitating accurate and timely detection methods for effective treatment. Deep learning-based approaches have emerged as promising solutions for automated medical image analysis. This study proposes an enhanced Mask R-CNN framework tailored specifically for the automatic detection and severity analysis of lung cancer from CT images. The proposed approach consists of three stages namely, pre-processing, lung nodule detection, and segmentation using enhanced Mask R-CNN and severity analysis. Our framework employs a deep convolutional neural network architecture trained on a comprehensive dataset of annotated lung images. By incorporating a region-based convolutional neural network (R-CNN) with a mask prediction branch, our model accurately localizes lung tumors while providing precise pixel-level segmentation masks. To enhance the performance of mask RCNN, the parameter present in the classifier is optimally selected using the adaptive pelican optimization (APO) algorithm. The proposed framework detects lung tumors and provides a comprehensive severity analysis, enabling clinicians to assess cancer stage and progression accurately. Evaluation on a benchmark dataset demonstrates superior detection accuracy and robustness compared to existing methods. Our enhanced Mask R-CNN approach shows promise as a valuable tool for early diagnosis and severity assessment of lung cancer, potentially improving patient outcomes and healthcare efficiency. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Pedestrian Trajectory Prediction with Missing Data: Datasets, Imputation, and Benchmarking 38

Pedestrian Trajectory Prediction with Missing Data: Datasets...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Chib, Pranav Singh Singh, Pravendra Department of Computer Science and Engineering Indian Institute of Technology Roorkee India

Pedestrian trajectory prediction is crucial for several applications such as robotics and self-driving vehicles. Significant progress has been made in the past decade thanks to the availability of pedestrian trajectory datasets, which enable trajectory prediction methods to learn from pedestrians' past movements and predict future trajectories. However, these datasets and methods typically assume that the observed trajectory sequence is complete, ignoring real-world issues such as sensor failure, occlusion, and limited fields of view that can result in missing values in observed trajectories. To address this challenge, we present TrajImpute, a pedestrian trajectory prediction dataset that simulates missing coordinates in the observed trajectory, enhancing real-world applicability. TrajImpute maintains a uniform distribution of missing data within the observed trajectories. In this work, we comprehensively examine several imputation methods to reconstruct the missing coordinates and benchmark them for imputing pedestrian trajectories. Furthermore, we provide a thorough analysis of recent trajectory prediction methods and evaluate the performance of these models on the imputed trajectories. Our experimental evaluation of the imputation and trajectory prediction methods offers several valuable insights. Our dataset provides a foundational resource for future research on imputation-aware pedestrian trajectory prediction, potentially accelerating the deployment of these methods in real-world applications. Publicly accessible links to the datasets and code files are available at https://***/Pranav-chib/TrajImpute. © 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

Bridging the Knowledge Gap: Improving BERT models for answering MCQs by using Ontology-generated synthetic MCQA Dataset 37

Bridging the Knowledge Gap: Improving BERT models for answer...

引用

37th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2024

作者： Sahil, Sahil Kumar, P. Sreenivasa Department of Computer Science and Engineering Indian Institute of Technology Madras India

BERT-based models possess impressive language understanding capabilities but often lack domain-specific knowledge, limiting their performance on specialised tasks such as medical multiple-choice question answering (MCQA). In this paper, we study how biomedical ontologies, rich repositories of medical knowledge, can be harnessed to enhance BERT-based models for medical MCQA task. Our contributions include OntoMCQA-Gen, a system which leverages different biomedical ontologies to construct BioOntoMCQA, a large synthetic MCQA dataset. OntoMCQA-Gen exploits the subclass-class relationships, definitions of concepts, and also synonym relationships from the ontologies to create this dataset of MCQs automatically. We then use this synthetic dataset to fine-tune various BERT-based models to answer medical MCQs. We evaluated these fine-tuned BERT models on the challenging MedMCQA and MedQA datasets of questions from admission examinations for medical degrees in India and USA, respectively. Our evaluation study on these datasets shows that fine-tuning the BERT-based models on BioOntoMCQA results in significantly improved accuracy scores. BioBERT and PubMedBERT, pretrained on the large medical corpus, have also shown significant improvements with our technique of fine-tuning ontology-generated synthetic data. This finding highlights the effectiveness of incorporating biomedical ontologies to enhance the BERT-based model in the medical domain. Moreover, our results underscore the importance of using ontology-generated data along with model adaptation for specialised domains, contributing to a novel advancement in natural language processing. Copyright © 2024 by the authors.

关键词： Ontology

来源：评论

学校读者我要写书评

暂无评论

Remember This Event That Year? Assessing Temporal Information and Understanding in Large Language Models

Remember This Event That Year? Assessing Temporal Informatio...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Beniwal, Himanshu Patel, Dishant Kowsik Nandagopan, D. Ladia, Hritik Yadav, Ankit Singh, Mayank Department of Computer Science and Engineering Indian Institute of Technology Gandhinagar India

ISBN: (纸本)9798891761681

Large Language Models (LLMs) are increasingly ubiquitous, yet their ability to retain and reason about temporal information remains limited, hindering their application in real-world scenarios where understanding the sequential nature of events is crucial. Our study experiments with 12 state-of-the-art models (ranging from 2B to 70B+ parameters) on a novel numerical-temporal dataset, TempUN, spanning from 10,000 BCE to 2100 CE, to uncover significant temporal retention and comprehension limitations. We propose six metrics to assess three learning paradigms to enhance temporal knowledge acquisition. Our findings reveal that open-source models exhibit knowledge gaps more frequently, suggesting a trade-off between limited knowledge and incorrect responses. Additionally, various fine-tuning approaches significantly improved performance, reducing incorrect outputs and impacting the identification of 'information not available' in the generations. The associated dataset and code are available at https://***/lingoiitgn/TempUN. © 2024 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

Fastlane: A framework for building fast path network applications

Fastlane: A framework for building fast path network applica...

引用

2024 IEEE Conference on Network Function Virtualization and Software Defined Networks, NFV-SDN 2024

作者： Das, Debojeet Lakshya Vutukuru, Mythili Department of Computer Science and Engineering Indian Institute of Technology Bombay India

ISBN: (纸本)9798350380538

To leverage the full potential of high-speed network interface cards (NICs), it is essential for network applications to use high-performance packet I/O frameworks. However, porting applications to use these specialized frameworks is a tedious task. While prior work has proposed frameworks to ease the development of network functions over fast packet I/O mechanisms, there has not been much work done to ease the porting of applications that are built to run on the kernel network stack. This paper presents Fastlane, a framework to easily port the "fast path" of traditional network applications to use fast packet I/O mechanisms while retaining the "slow path" to run on the traditional network stack. Fastlane provides APIs by which application developers can configure, use, and switch between different fast packet I/O mechanisms like DPDK and CNDP in the application’s fast path. Additionally, Fastlane provides APIs for the fast path to communicate with the slow path running on the kernel network stack. Fastlane also allows applications to be deployed in an unprivileged Kubernetes pod, easing application development for cloud deployments. We evaluate the effectiveness of Fastlane by porting a production-grade 5G user plane function to use our framework. Our evaluation shows that Fastlane enables developers to build high-performance network applications easily and imposes minimal performance overheads. ©2024 IEEE.

关键词： HIgh speed networks

来源：评论

学校读者我要写书评

暂无评论

A Systematic Study for Understanding the Security Risks in 5G Core Network 16

A Systematic Study for Understanding the Security Risks in 5...

引用

16th International Conference on COMmunication Systems and NETworkS, COMSNETS 2024

作者： Dixit, Unnati Vittal, Shwetha Antony Franklin, A. Indian Institute of Technology Hyderabad Department of Computer Science and Engineering India

ISBN: (纸本)9798350383119

The widespread adoption of 5G networks has revolutionized communication, providing great connectivity and services across various verticals. Specifically, on the 5G Core (5GC) this has been made possible by leveraging network slicing and Network Function Virtualization (NFV) where multiple Network Functions (NFs) are involved to serve different user requests on the control plane and a variety of service traffic on the data plane. However, as 5G networks embrace zero trust and multi-tenant architectures, this rapid technological advancement has also exposed the 5GC to various security risks such as Distributed Denial of Service (DDoS), Man in The Middle (MITM), cross-slice disruptions, and side-channel attacks. In this regard, this paper presents a comprehensive approach to enhance the security of 5GC through a systematic study of dependency graphs, Betweenness Centrality (BC), and attack graphs. Potential attack scenarios are carefully analyzed by modeling the 3GPP-defined 5GC, performing vulnerability assessments, building attack graphs, analyzing them, and presenting the key inferences from the study. © 2024 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Parallel Algorithms For Exact SimRank Computations 31

Efficient Parallel Algorithms For Exact SimRank Computations

引用

31st IEEE International Conference on High Performance Computing, Data, and Analytics Workshops, HiPCW 2024

作者： Mundhara, Aditya Nijhara, Prajjwal Banerjee, Dip Sankar Department of Computer Science and Engineering Indian Institute of Technology Jodhpur India

ISBN: (纸本)9798331509118

SimRank is a popular measure for evaluating node similarities in graphs, but its high computational cost limits scalability for large graphs. The ExactSim [1] algorithm achieves precise single-source SimRank similarities but suffers from significant computation time. In this work, we parallelize key components of ExactSim - Personalized PageRank (PPR) computations and SimRank calculations on a CPU-based shared memory multi-core system. By optimizing these modules, we significantly reduce runtime, making ExactSim more practical for large-scale applications. Our results (on graph sizes ranging from 0.1 million edges to 100 million edges) demonstrate the effectiveness of parallelization in improving both efficiency and scalability. © 2024 IEEE

关键词： Graph algorithms

来源：评论

学校读者我要写书评

暂无评论

Cross-lingual Editing in Multilingual Language Models 18

Cross-lingual Editing in Multilingual Language Models

引用

18th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2024 - Findings of EACL 2024

作者： Beniwal, Himanshu Kowsik Nandagopan, D. Singh, Mayank Department of Computer Science and Engineering Indian Institute of Technology Gandhinagar India

ISBN: (纸本)9798891760936

The training of large language models (LLMs) necessitates substantial data and computational resources, and updating outdated LLMs entails significant efforts and resources. While numerous model editing techniques (METs) have emerged to efficiently update model outputs without retraining, their effectiveness in multilingual LLMs, where knowledge is stored in diverse languages, remains an underexplored research area. This research paper introduces the cross-lingual model editing (XME) paradigm, wherein a fact is edited in one language, and the subsequent update propagation is observed across other languages. To investigate the XME paradigm, we conducted experiments using BLOOM, mBERT, and XLM-RoBERTa using the two writing scripts: Latin (English, French, and Spanish) and Indic (Hindi, Gujarati, and Bengali). The results reveal notable performance limitations of state-of-the-art METs under the XME setting, mainly when the languages involved belong to two distinct script families. These findings highlight the need for further research and development of XME techniques to address these challenges. For more comprehensive information, the dataset used in this research and the associated code are publicly available at the following URL. © 2024 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

PCLive: Pipelined Restoration of Application Containers for Reduced Service Downtime 15

PCLive: Pipelined Restoration of Application Containers for ...

引用

15th Annual ACM Symposium on Cloud Computing, SoCC 2024

作者： Tripathi, Shiv Bhushan Mishra, Debadatta Department of Computer Science and Engineering Indian Institute of Technology Kanpur India

ISBN: (纸本)9798400712869

Application containers are widely used in contemporary cloud computing environments. Migration of containers across hosts provides cost-effective cloud management by enabling improved server consolidation, load balancing and enhanced fault tolerance. One of the primary objectives of container migration is to reduce the service downtime of applications hosted in containers. The service downtime depends on performing the migration activities efficiently, specifically from the time the container is stopped on the source host till it is restored and fully functional at the destination host. In this paper, we show that, the state-of-the-art pre-copy migration strategy for containers using checkpoint and restore techniques (e.g., CRIU) inflates the downtime due to its inherent limitations in the restoration procedures, particularly for containers with large memory working set size. We propose PCLive to address this bottleneck using a pipelined restore mechanism. Compared to the baseline CRIU pre-copy migration, PCLive results in up to ∼38.8x reduction in restoration time which leads to a reduction of service downtime by up to ∼2.7x for migration of a container hosting the Redis key-value store over an one Gbps network. We also present comprehensive comparative analysis of the resource cost for the proposed solution along with additional optimizations to demonstrate that PCLive can reduce the application downtime in a resource efficient manner leveraging its flexible and efficient design choices. © 2024 Owner/Author.

关键词： Restoration

来源：评论

学校读者我要写书评

暂无评论

Impact of Constraint Modifications on the CRYSTALS-Kyber Hardware Design 31

Impact of Constraint Modifications on the CRYSTALS-Kyber Har...

引用

31st IEEE International Conference on High Performance Computing, Data, and Analytics Workshops, HiPCW 2024

作者： Bohra, Dixit Dutt Banerjee, Dip Sankar Sanadhya, Somitra Department of Computer Science and Engineering Indian Institute of Technology Jodhpur India

ISBN: (纸本)9798331509118

The lattice-based post-quantum cryptography scheme is widely regarded as one of the most secure solutions to address the emerging threats posed by advancements in quantum computing. Various groups have analyzed and studied the hardware implementation of the CRYSTALS-KYBER design, and they continue to work on optimizing the design. Researchers have applied various techniques throughout the design to enhance power consumption, area utilization, and output speed. The main goal of this work is to present the impact of constraints on design performance, analyzing factors such as Worst Negative Slack (WNS) and total power. After introducing input and output delay constraints, we found that the actual operating frequency was 91.29 MHz, deviating significantly from the initially claimed 169.92 MHz. © 2024 IEEE

关键词： Quantum computers

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：