检索结果-内蒙古大学图书馆

International Conference on Inventive Computation Technologies (ICICT)

作者： Shajulin Benedict Rubiya Subair Tanya Gupta Vedanta S.P. Dept. of Computer Science and Engineering Indian Institute of Information Technology Kottayam Kerala India

The step to the success of startups is through overcoming competitors by adopting software innovations that improve businesses. Serverless computing model, recently, has intrigued a sizable number of startup professionals belonging to various sectors, including financial or IoT-enabled application developers. One of the main flaws is its heavy dependency on cloud providers, which can still result in hefty pricing to startups and stalling functions in applications. This article proposes a penaltyenabled serverless architecture for startups. The architecture can boost the economy of startups and can analyze the serverlessoriented cost-saving options in applications. The penalty-oriented approach could enable cloud architects, developers, and startups, to rethink the utilization of serverless functions; to gleam of with future innovations.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Device Identification using Network Traces in a Smart Home IoT Network

Device Identification using Network Traces in a Smart Home I...

引用

International Symposium on Advanced Networks and Telecommunication Systems (ANTS)

作者： Jwngfu Brahma Rakesh Matam Nikumani Choudhury Dept. of Computer Science & Engineering Indian Institute of Information Technology Guwahati India Dept. of Computer Science & Information Systems Birla Institute of Technology & Science Hyderabad India

Device identification is a crucial aspect of securing networks, particularly in the context of the Internet of Things (IoT), where a vast variety of devices are interconnected. Recently, there has been significant research on developing techniques for identifying IoT devices based on their unique characteristics, called as device fingerprints. These techniques use machine learning algorithms, which can effectively learn and classify devices based on their features. However, device identification remains a challenging task due to the diversity of IoT devices and the constant evolution of their characteristics. This paper presents our attempt at device identification by analyzing distinct device characteristics observed during network communication and using different machine learning techniques.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Prediction of Parkinson's Disease using Hybrid Feature Selection based Techniques 4

Prediction of Parkinson's Disease using Hybrid Feature Selec...

引用

4th International Conference on Computing and Communications Technologies, ICCCT 2021

作者： Dash, Avijit Kumar Indian Institute of Technology Bombay Dept. of Computer Science and Engineering Powai India

ISBN: (纸本)9781665414470

Parkinson's disease (PD) is one of the significant severe problems globally in recent times. It is a neurological disorder that progresses over time and the most severe problems after Alzheimer's disease. Our article proposes a Hybrid Feature Selection system for the initial detection of PD from speech recordings. This method picks the best set of instances that can lessen instance vector dimensions from 22 to 5. We have proposed a machine learning-based model using five different classifiers named Random Forest, Logistic Regression, XGBoost, AdaBoost, and Gradient Tree Boosting. Gradient Tree Boosting presents the best appearance with a spectacular accuracy of 98.31% and the area under the ROC curve 98.66%, among all classifiers used to predict PD. We showed that the stated design has greater accuracy than the current methods available in the literature, and the number of instances is less than others. © 2021 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

LIMMITS’24: Multi-Speaker, Multi-Lingual Indic TTS with Voice Cloning

LIMMITS’24: Multi-Speaker, Multi-Lingual Indic TTS with Voi...

引用

Acoustics, Speech, and Signal Processing Workshops (ICASSPW), IEEE International Conference on

作者： Abhayjeet Singh Amala Nagireddi Deekshitha G Jesuraja Bandekar Roopa R Sandhya Badiger Sathvik Udupa Prasanta Kumar Ghosh Hema A Murthy Pranaw Kumar Keiichi Tokuda Mark Hasegawa-Johnson Philipp Olbrich Electrical Engineering Department Indian Institute of Science (IISc) Bangalore India Dept of Computer Science and Engineering Indian Institute of Technology Madras India CDAC Mumbai India Department of Computer Science Nagoya Institute of Technology Japan Department of Electrical and Computer Engineering University of Illinois Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH Bonn Germany

ISBN: (数字)9798350374513

ISBN: (纸本)9798350374520

The Multi-speaker, Multi-lingual Indic TTS with voice cloning (LIMMITS’24) challenge is organized as part of the ICASSP 2024 signal processing grand challenge. LIMMITS’24 aims at the development of voice cloning for multi-speaker, multi-lingual Text-to-Speech (TTS) model. Towards this, 80 hours of TTS data has been released in each of Bengali, Chhattisgarhi, English (indian), and Kannada languages. This is in addition to Telugu, Hindi, and Marathi data released in the LIMMITS’23. The challenge encourages the advancement of TTS in indian Languages as well as the development of multi-speaker voice cloning techniques for TTS. The three tracks of LIMMITS’24 have provided an opportunity for various researchers and practitioners around the world to explore the state of the art in TTS research.

关键词： Conferences Cloning Signal processing Acoustics Text to speech Speech processing

来源：评论

学校读者我要写书评

暂无评论

Graph Convolutional Network-based Centre-Object Tracking for the multi-dynamic agents for Autonomous Vehicles 5

Graph Convolutional Network-based Centre-Object Tracking for...

引用

5th International Conference on Information Systems and computer Networks, ISCON 2021

作者： Singh, Divya Srivastava, Rajeev Indian Institute of Technology BHU Dept. of Computer Science and Engineering Varanasi India

ISBN: (纸本)9781665403412

Multi-Object tracking goals to localize, classify and track all object instances of each class throughout an image sequence. It is very useful to understand the video scenes and very desirable for computer vision-based applications that require detailed pixel-level information, such as video editing and autonomous driving. Our method involves Detection-based tracking, which generally requires detected objects with a tracker that links the boundary boxes (bboxes) to generate target trajectory. To optimize the similarity between tracked and detected objects are the basic part of multi-object tracking (MOT) frameworks. To mark this issue, ongoing works concurrently enhance the matching technique of data modules with joint MOT structure, which has expressed improved execution. The key thought behind our GCNN approach is to take into consideration and learn the distinctive or selective features for local and global matching. It takes into account the relations between nearby objects in a frame sequence and thus provides a better tracking performance as compared to the existing models. We demonstrated our model on the MOT datasets that showed the viability of our GCNN-based joint MOT approach which includes the Hungarian algorithm to assign the identity of the object in the next frame. © 2021 IEEE.

关键词： Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Federated matched averaging with information-gain based parameter sampling 2021

Federated matched averaging with information-gain based para...

引用

1st International Conference on AI-ML-Systems, AIMLSystems 2021

作者： Shukla, Sharad Srivastava, Nisheeth Dept. of Computer Science and Engineering Indian Institute of Technology Kanpur Kanpur India

ISBN: (纸本)9781450385947

Federated learning allows edge devices to learn a shared global model from the client's model parameters while keeping the training data on the device. However, for large models, transmitting all model parameters imposes considerable communication costs, which can be a significant bottleneck in bandwidth-constrained deployments. We present a federated matched averaging algorithm with information-gain based sampling that considerably reduces the number of parameters to be sent by all clients in a federated learning paradigm. Experiments across five benchmark datasets, encompassing symbolic, image and audio data, suggest that our algorithm significantly reduces communication overhead in federated learning settings, without reducing classification accuracy. © 2021 ACM.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Integrating Cloud And Edge Computing For Scalable Iot Applications 2

Integrating Cloud And Edge Computing For Scalable Iot Applic...

引用

2nd International Conference on Advances in Computation, Communication and Information technology, ICAICCIT 2024

作者： Nagarjuna, Pitty Sevuga Pandian, A. Sharma, Suresh Kumar Ramesh, T. Sohail, Mohammed Ali Dinesh Babu, M. Indian Institute of Science Karnataka Bengaluru India Department of Computer Science Kristu Jayanti College Karnataka Bengaluru India Dept. of Statistics Mathematics and Computer Science Sri Karan Narendra Agriculture University Jaipur Jobner India Department of CSE R.M.K Engineering College Chennai India Department of EEE Jazan University Jazan Saudi Arabia Department of Mechanical Rajalakshmi Institute of Technology Chennai India

ISBN: (纸本)9798331541217

The data from the Internet of Things (IoT) is essential in the modern data-driven digital economy since it inspires many new business models supplying a wide spectrum of services both ubiquitous and intelligent. The data transmitted using edge computing technologies is investigated in this paper. Under the framework of the trust model of the research, the data exchanged between the source block and the destination block generates direct, indirect, and reciprocal trust. Although cloud computing can serve as a paradigm for safe data flow, the results of the study revealed that a trust-based approach can avoid cloud intrusion. The simulation runs on numerous test beds using the presently in use cloud technique. The results clearly illustrate that the trust-based approach is effective in protecting data moved across edge computing. © 2024 IEEE.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

SMART: Submodular Data Mixture Strategy for Instruction Tuning

arXiv

引用

arXiv 2024年

作者： Renduchintala, H.S.V.N.S. Kowndinya Bhatia, Sumit Ramakrishnan, Ganesh Media and Data Science Research Adobe Inc. India Dept. of Computer Science & Engineering Indian Institute of Technology Bombay India

Instruction Tuning involves finetuning a language model on a collection of instruction-formatted datasets in order to enhance the generalizability of the model to unseen tasks. Studies have shown the importance of balancing different task proportions during finetuning, but finding the right balance remains challenging. Unfortunately, there’s currently no systematic method beyond manual tuning or relying on practitioners’ intuition. In this paper, we introduce SMART (Submodular data Mixture strAtegy for instRuction Tuning) — a novel data mixture strategy which makes use of a submodular function to assign importance scores to tasks which are then used to determine the mixture weights. Given a fine-tuning budget, SMART redistributes the budget among tasks and selects non-redundant samples from each task. Experimental results demonstrate that SMART significantly outperforms traditional methods such as examples proportional mixing and equal mixing. Furthermore, SMART facilitates the creation of data mixtures based on a few representative subsets of tasks alone and through task pruning analysis, we reveal that in a limited budget setting, allocating budget among a subset of representative tasks yields superior performance compared to distributing the budget among all tasks. The code for reproducing our results is open-sourced at https://***/kowndinyarenduchintala/SMART. © 2024, CC BY.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Integer Wavelet Transform based Secured Image Steganography using LSB and Coefficient Value Differencing 2

Integer Wavelet Transform based Secured Image Steganography ...

引用

2nd International Conference on Secure Cyber Computing and Communications, ICSCCC 2021

作者： Mandal, Pratap Chandra Mukherjee, Imon Indian Institute of Information Technology Kalyani B.P. Poddar Institute of Management and Technology Dept. of Computer Science and Engineering Kolkata India Indian Institute of Information Technology Kalyani Dept. of Computer Science and Engineering India

ISBN: (纸本)9781665444156

A steganography technique based on the Integer Wavelet Transform (IWT) is proposed. High-frequency coefficients are exploited more to embed the confidential message since they are less sensitive to the human eye. Based on this notion, the 2-LSB method is employed in the wavelet coefficients of the approximation subband which contains most of the signal energy and the 4-LSB method is applied in the wavelet coefficients of the diagonal subband. The Coefficient Value Differencing (CVD) method has been introduced in the horizontal and vertical subbands. To obtain the highest embedding capacity with an acceptable distortion, a threshold value has been applied in the CVD method. The proposed technique achieves an average embedding capacity of 2.28 bpp for all the stego images while maintaining a lower distortion. It shows strength against the standard tool, StirMark Benchmark 4.0, and withstands the steganalysis attacks. © 2021 IEEE.

关键词： Steganography

来源：评论

学校读者我要写书评

暂无评论

Robust Facial Emotion Recognition System via De-Pooling Feature Enhancement and Weighted Exponential Moving Average

Robust Facial Emotion Recognition System via De-Pooling Feat...

引用

IEEE Region 10 International Conference TENCON

作者： Khin Cho Win Zahid Akhtar C. Krishna Mohan Dept. of Artificial Intelligence Indian Institute of Technology Hyderabad India Dept. of Network and Computer Security State University of New York Polytechnic Institute Utica NY USA Dept. of Computer Science and Engineering Indian Institute of Technology Hyderabad India

ISBN: (数字)9798350350821

ISBN: (纸本)9798350350838

Facial expression recognition (FER) in an uncontrolled environment presents a formidable challenge in affective computing and human-machine interaction domains. Existing FER models fail to generalize due to the innate expression nature of the intra-class separability and inter-class compactness. Although most state-of-the-art FER research focuses on more advanced frameworks and well-separated discriminative loss functions, further exploration is needed to obtain high-quality expression features. This paper proposes a robust feature enhancement approach for FER by integrating the De-pooling Feature Enhancement and Weighted Exponential Moving Average (WEMA) of Stochastic Gradient Descent (SGD). The proposed method utilizes De-Pooling Feature Enhancement to capture and enhance the high-quality expression features, while WEMA of SGD optimizes the training process for improved stability and convergence. Our extensive experimental analysis of benchmark datasets signifies that our proposed method prevails over state-of-the-art methods, achieving superior accuracy performances of 86.43% on FER2013, 94.97% on FERPlus, 94.71% on RAF-DB, as well as 80.62% and 72.90% on AffectNet, of 7 class and 8 class, respectively.

关键词： Training Emotion recognition Face recognition Computational modeling Neural networks Stochastic processes Stability analysis Robustness IEEE Regions Convergence

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：