检索结果-内蒙古大学图书馆

Real-time earthquake magnitude prediction using designed machine learning ensemble trained on real and CTGAN generated synthetic data

引用

Geodesy and Geodynamics 2025年第3期16卷 350-368页

作者： Anushka Joshi Balasubramanian Raman C.Krishna Mohan Department of Computer Science and Engineering Indian Institute of Technology

The earthquake early warning(EEW) system provides advance notice of potentially damaging ground shaking. In EEW, early estimation of magnitude is crucial for timely rescue operations. A set of thirty-four features is extracted using the primary wave earthquake precursor signal and site-specific *** Japan's earthquake magnitude dataset, there is a chance of a high imbalance concerning the earthquakes above strong impact. This imbalance causes a high prediction error while training advanced machine learning or deep learning models. In this work, Conditional Tabular Generative Adversarial Networks(CTGAN), a deep machine learning tool, is utilized to learn the characteristics of the first arrival of earthquake P-waves and generate a synthetic dataset based on this information. The result obtained using actual and mixed(synthetic and actual) datasets will be used for training the stacked ensemble magnitude prediction model, MagPred, designed specifically for this study. There are 13295, 3989, and1710 records designated for training, testing, and validation. The mean absolute error of the test dataset for single station magnitude detection using early three, four, and five seconds of P wave are 0.41, 0.40,and 0.38 MJMA. The study demonstrates that the Generative Adversarial Networks(GANs) can provide a good result for single-station magnitude prediction. The study can be effective where less seismic data is available. The study shows that the machine learning method yields better magnitude detection results compared with the several regression models. The multi-station magnitude prediction study has been conducted on prominent Osaka, Off Fukushima, and Kumamoto earthquakes. Furthermore, to validate the performance of the model, an inter-region study has been performed on the earthquakes of the India or Nepal region. The study demonstrates that GANs can discover effective magnitude estimation compared with non-GAN-based methods. This has a high potential for wid

关键词： Magnitude Synthetic data Machine learning Earthquake CTGAN

来源：评论

学校读者我要写书评

暂无评论

Sentiment and hashtag-aware attentive deep neural network for multimodal post popularity prediction

引用

Neural Computing and Applications 2024年第4期37卷 2799-2824页

作者： Bansal, Shubhi Kumar, Mohit Raghaw, Chandravardhan Singh Kumar, Nagendra Department of Computer Science and Engineering Indian Institute of Technology Indore Khandwa Road Simrol Madhya Pradesh Indore453552 India

Social media users articulate their opinions on a broad spectrum of subjects and share their experiences through posts comprising multiple modes of expression, leading to a notable surge in such multimodal content on social media platforms. Nonetheless, accurately forecasting the popularity of these posts presents a considerable challenge. Prevailing methodologies primarily center on the content itself, thereby overlooking the wealth of information encapsulated within alternative modalities such as visual demographics, sentiments conveyed through hashtags and adequately modeling the intricate relationships among hashtags, texts, and accompanying images. This oversight limits the ability to capture emotional connection and audience relevance, significantly influencing post popularity. To address these limitations, we propose a seNtiment and hAshtag-aware attentive deep neuRal netwoRk for multimodAl posT pOpularity pRediction, herein referred to as NARRATOR that extracts visual demographics from faces appearing in images and discerns sentiment from hashtag usage, providing a more comprehensive understanding of factors influencing post popularity. Moreover, we introduce a hashtag-guided attention mechanism that leverages hashtags as navigational cues, guiding the model’s focus toward the most pertinent features of textual and visual modalities, thus aligning with target audience interests and broader social media context. Experimental results demonstrate that NARRATOR outperforms existing methods by a significant margin on two real-world datasets. Furthermore, ablation studies underscore the efficacy of integrating visual demographics, sentiment analysis of hashtags, and hashtag-guided attention mechanisms in enhancing the performance of post popularity prediction, thereby facilitating increased audience relevance, emotional engagement, and aesthetic appeal. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Demonstration of an Application Layer Omni-frequency IoT Protocol 17

Demonstration of an Application Layer Omni-frequency IoT Pro...

引用

17th International Conference on COMmunication Systems and NETworkS, COMSNETS 2025

作者： Gupta, Nayancy Peddamalla, Lokesh Mendu, Veman Sai Prabhath Banda, Gourinath Indian Institute of Technology Indore Department of Computer Science and Engineering Indore India

ISBN: (纸本)9798331531195

The advent of the Internet of Things (IoT) has transformed the way devices communicate, with an ever-increasing need for seamless interoperability and energy-efficient communication. This paper presents a unified omni-frequency protocol designed to operate independently of specific radio-wave protocols, allowing efficient and secure communication between users, devices (referred to as "things"), and cloud servers. The proposed protocol operates at the Application Layer and introduces two key components: the original equipment manufacturer (OEM) Server, which centrally manages user and device data, and the OEM Host, which verifies device credentials. This system enables efficient Internet Protocol(IP)-based and non-IP based communication, improving energy management and distribution of computational tasks to the edge. An Android application was developed to serve as the user interface, supporting multiple communication methods, including IP, Bluetooth classic version 5.0, Near-Field Communication (NFC) and ad hoc networking. The versatility of this unified approach optimizes the interactions, data processing, and energy utilization of IoT in various practical environments. © 2025 IEEE.

关键词： Network layers

来源：评论

学校读者我要写书评

暂无评论

HLS driven Hybrid GA-PSO for Design Space Exploration of Optimal Palmprint Biometric based IP Watermark and Loop Unrolling Factor 10

HLS driven Hybrid GA-PSO for Design Space Exploration of Opt...

引用

10th IEEE International Symposium on Smart Electronic Systems, iSES 2024

作者： Sengupta, Anirban Chourasia, Vishal Kumar, Nitish Computer Science and Engineering Indian Institute of Technology Indore Indore India

ISBN: (纸本)9798331533229

This paper proposes a novel high level synthesis (HLS) based hybrid Genetic Algorithm-Particle Swarm Optimization (GA-PSO) framework for concurrent design space exploration (DSE) of optimal palmprint biometric based intellectual property (IP) watermark and optimal loop unrolling factor. The proposed approach is a hybrid exploration framework that merges GA process with specific utility attributes of PSO based DSE framework such as mutation, local and global best design solution concepts, which aids in escaping getting stuck at local minima. The proposed approach is capable of providing low-cost robust detective security countermeasure against threats of IP ownership abuse and IP piracy using IP seller's optimal palmprint biometric watermark. The comparison of the proposed approach with related works demonstrated optimal quality palmprint watermark in terms of lower probability of coincidence (watermark collision) and stronger tamper tolerance ability, at zero design cost overhead. © 2024 IEEE.

关键词： Space research

来源：评论

学校读者我要写书评

暂无评论

HLS Based Rapid Pareto Front Search of Watermarked Convolutional Layer IP Design 10

HLS Based Rapid Pareto Front Search of Watermarked Convoluti...

引用

10th IEEE International Symposium on Smart Electronic Systems, iSES 2024

作者： Sengupta, Anirban Chourasia, Vishal Computer Science and Engineering Indian Institute of Technology Indore Indore India

ISBN: (纸本)9798331533229

This paper presents a novel methodology for high level synthesis (HLS) based Pareto front search of watermarked convolutional layer intellectual property (IP) designs. The presented exploration framework is rapidly capable of pruning the design space of watermarked IP designs using a combination of architecture tree configuration and dominance factor logic. Experimental results validate that the proposed approach is capable of providing significant speedup in the exploration time compared to exhaustive analysis. Further, results of security analysis in terms of watermark robustness achieved by the proposed approach against standard attacks of false ownership claim, tampering and watermark collision have also been reported. © 2024 IEEE.

关键词： High level synthesis

来源：评论

学校读者我要写书评

暂无评论

HLS Based Hardware Watermarking Using IP Seller's Superimposed Facial Anthropometric Features 10

HLS Based Hardware Watermarking Using IP Seller's Superimpos...

引用

10th IEEE International Symposium on Smart Electronic Systems, iSES 2024

作者： Sengupta, Anirban Anshul, Aditya Chourasia, Vishal Computer Science and Engineering Indian Institute of Technology Indore Indore India

ISBN: (纸本)9798331533229

Ensuring the security of hardware intellectual property (IP) designs is crucial for IP sellers/vendors due to the exposure of the global design chain process to security risks like IP piracy and false IP ownership claim. A potential threat within system-on-chip (SoC) or foundry house may involve piracy or fraudulent claim of ownership rights by an attacker. Therefore, ensuring the security of IP designs is paramount from an IP seller's perspective. This paper presents a novel high level synthesis (HLS) based hardware watermarking by exploiting IP seller's superimposed facial anthropometric features as imperceptible embedded watermark. The watermark is embedded into the hardware IP design during HLS as covert security constraints which serves as digital evidence against external threats. The proposed approach offers enhanced security (i.e., stronger tamper tolerance and lower probability of coincidence) compared to prior works at negligible design cost overhead. © 2024 IEEE.

关键词： System-on-chip

来源：评论

学校读者我要写书评

暂无评论

TRIESTE: translation based defense for text classifiers

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第12期14卷 16385-16396页

作者： Gupta, Anup Kumar Paliwal, Vardhan Rastogi, Aryan Gupta, Puneet Department of Computer Science and Engineering Indian Institute of Technology Indore Madhya Pradesh Indore452020 India Department of Electrical Engineering Indian Institute of Technology Indore Madhya Pradesh Indore452020 India

The field of natural language processing (NLP) has significantly evolved with the advent of state-of-the-art models. The discovery of these models has entirely revolutionised how NLP tasks such as machine translation, sentiment analysis and many others are performed. However, despite their high efficacy and meticulous performance, these models are prone to adversarial attacks. Adversarial attacks involve the introduction of perturbations imperceptible to humans, which can severely impact the model’s learning and prediction accuracy. Current defenses on text data include approaches such as spell-checking and adversarial training, which have their limitations against state-of-the-art adversarial attacks. This paper put forward an effective transformation-based defense, TRIESTE (TRanslatIon basEd defenSe for Text classifiErs). The proposed defense overcomes the shortcomings of existing defenses by translating the input text from the source language to a target language and again back to the source language before providing it to the text classifier. Translation ensures that the sentiment of the translated text is similar to that of the input text by taking the entire text into consideration, which leads to the removal of adversarial perturbations. Rigorous evaluation on publicly available datasets showcases that TRIESTE is successful against state-of-the-art attacks without a significant drop in the classifier accuracy. © 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Gen-Sign: HLS Based Watermarking Using IP Vendor's Feistel Cipher Encrypted Genomic Signature for Protecting CNN and Image Processing Filter Cores Against Piracy 10

Gen-Sign: HLS Based Watermarking Using IP Vendor's Feistel C...

引用

10th IEEE International Symposium on Smart Electronic Systems, iSES 2024

作者： Sengupta, Anirban Chourasia, Vishal Singh, Ayush Kumar Computer Science and Engineering Indian Institute of Technology Indore Indore India

ISBN: (纸本)9798331533229

This paper presents a novel hardware intellectual property (IP) watermarking process during high level synthesis (HLS) that leverages IP vendor human body sample in the form of genomic sequence to act as secret invisible encrypted signature (called 'Gen-Sign'), The proposed encrypted genomic signature is used as an imperceptible watermark to secure convolutional neural network (CNN) and image processing filter against IP piracy. The presented approach has been evaluated on benchmarks in terms of design cost overhead, probability of coincidence, tamper tolerance and entropy. Results of the proposed approach indicate superior security at 0% design overhead, compared to prior approaches. © 2024 IEEE.

关键词： Crime

来源：评论

学校读者我要写书评

暂无评论

District-Level Rainfall and Cloudburst Prediction Using XGBoost: A Machine Learning Approach for Early Warning Systems

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第2期49卷 375-396页

作者： Kumar, Guru Dayal Tyagi, Shekhar Pradhan, Kalandi Charan Shah, Akshat Migration and Development Research Group School of Humanities and Social Sciences Indian Institute of Technology Indore452020 India Department of Computer Science and Engineering Indian Institute of Technology Indore452020 India Department of Computer Science and Engineering Shri Ramdeobaba College of Engineering and Management Nagpur440001 India

This research presents a novel methodology for cloudburst forecasting using the XGBoost (Extreme Gradient Boosting) machine learning algorithm. With the escalating impact of climate change, accurately predicting extreme weather events like cloudbursts is crucial due to their potential to trigger floods. Cloudburst events were identified from daily rainfall data. Our study leverages historical weather data, focusing on intricate rainfall patterns, to forecast future cloudburst occurrences. Comparative analysis against Random Forest and LSTM models confirmed XGBoost’s effectiveness, consistently outperforming alternatives across multiple performance metrics. The XGBoost model, known for its ability to handle complex datasets, demonstrated strong predictive performance, with an RMSE of 0.12 and an MAE of 0.09, indicating high accuracy. This research provides a reliable tool for advanced weather forecasting and early warning systems, offering valuable support to policymakers, disaster management teams, and agricultural planners in mitigating risks associated with extreme rainfall events. © 2025 Slovene Society Informatika. All rights reserved.

关键词： Acid rain

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Human Activity Recognition System for Assisted Living Environments

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第5期5卷 2342-2357页

作者： Jain, Ankit Akerkar, Rajendra Srivastava, Abhishek Indian Institute of Technology Indore Department of Computer Science and Engineering Madhya Pradesh Indore453552 India Western Norway Research Institute Sogndal6856 Norway

Automatic human activity recognition has numerous applications, especially in elderly support and healthcare. Several approaches for human activity recognition (HAR) using a variety of sensors are available in literature. While such frameworks are effective, each has limitations related to privacy, convenience, cost, and performance. In this article, a robust framework for automatic human activity recognition is proposed that uses depth sensors that preserve privacy and are cost-effective. The depth sensors provide two data modalities, namely, depth maps and skeleton sequences, used together for activity recognition. Two novel descriptors, joint position descriptor based on the position of joints;and bone angle descriptor based on bone inclination, are generated from the skeleton sequence data. The descriptors convey both spatial and temporal information and are scale and view-point invariant. Depth video clips are used along with the descriptors to deal with the issue of noisy and missing skeleton sequences. The data modalities and descriptors are fused using a two-level fusion strategy for a multichannel convolutional neural network (CNN) framework. The proposed system is validated and shown to be superior to the existing state of the art through comparisons over four widely used public datasets. A computational complexity analysis of the system confirms its efficacy in real time. A prototypical implementation of the proposed system further validates its practicability. © 2020 IEEE.

关键词： Neural networks

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：