Search

Your search keyword '"Karthikeyan S"' showing total 273 results
Start Over Author "Karthikeyan S" Publisher ieee
273 results on '"Karthikeyan S"'

39. AI-based content filtering system using an age prediction algorithm

Author

Karthikeyan. S and Ashutosh Upadhyay

Subjects
Artificial neural network, business.industry, Age prediction, Computer science, Deep learning, Machine learning, computer.software_genre, Facial recognition system, Task (project management), Identification (information), Robustness (computer science), Face (geometry), Artificial intelligence, business, computer
Abstract

Computer vision mainly focuses on the automatic extraction, analysis, and understanding of useful information from a single image or video. On the other hand, authenticity is emerging as one of the primary requirements in today's world by developing a system for computer vision complexity. Generally, two robust techniques such as age estimation and face recognition are required to maintain authenticity. In reality, fraud and scams are getting increased, so here this paper has proposed a new combined model for face recognition and age prediction. Face recognition has been implemented and presented in this paper by using a Deep Neural Network. The authenticity problem can be handled by using either facial recognition or age prediction alone; this study has presented a method that employs both of them together to enhance the system's robustness. So, first, this model detects the person's face, and then it predicts the person's age. If the individual is eligible to view the information or perform a task, their access will be limited; otherwise, their access will be restricted. So it helps to solve two difficulties in this case: the person's identification cannot be faked, and their age is also confirmed by the system. (CNN for the face, and mention technique for the age.)

Published
2021
Full Text
View/download PDF

49. Live Event Detection for People’s Safety Using NLP and Deep Learning

Author

Amrit Sen, Gayathri Rajakumaran, Miroslav Mahdal, Shola Usharani, Vezhavendhan Rajasekharan, Rajiv Vincent, and Karthikeyan Sugavanan

Subjects
Natural language processing (NLP), deep learning, audio, recording, CNN, LSTM, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

Today, humans pose the greatest threat to society by getting involved in robbery, assault, or homicide activities. Such circumstances threaten the people working alone at night in remote areas especially women. Any such kind of threat in real time is always associated with a sound/noise which may be used for an early detection. Numerous existing measures are available but none of them sounds efficient due to lack of accuracy, delays in exact prediction of threat. Hence a novel software-based prototype is developed to detect threats from a person’s surrounding sound/noise and automatically alert the registered contacts of victims by sending email, SMS, WhatsApp messages through their smartphones without any other hardware components. Audio signals from Kaggle dataset are visualized, analyzed using Exploratory Data Analytics (EDA) techniques. By feeding EDA outcomes into various Deep Learning models: Long short-term memory (LSTM), Convolutional Neural Networks (CNN) yields accuracy of 96.6% in classifying the audio-events.

Published
2024
Full Text
View/download PDF

50. EDgE: Enhancing Predictability in Decentralized Network Slices Through Federated Learning—A Multi Layered Ecosystem

Author

Ganesh Chandrasekaran, Abhinay Kumar, Karthikeyan Subramaniam, N. Sunil, and Rajesh Challa

Subjects
Network slicing, virtualization, management and orchestration, machine learning, federated learning, blockchain, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

Network slicing has emerged as a crucial technology in modern communication networks, especially in the context of 5G and beyond. Network slices offer virtualized and isolated segments tailored to the specific applications and services. However, the inherent isolation imposes security challenges, limiting the slice’s ability to learn from all data sources. This paper proposes EDgE, a novel solution leveraging federated learning (FL) to address these concerns. The proposed framework leverages FL to facilitate learning from diverse data sources, enabling each slice to conduct local machine learning (ML) model training with encrypted model updates shared among intra and inter-communication service providers (CSPs). The EDgE framework enables dynamic selection of parameters for learning and optimizing specific attributes based on the slice type and requirements. This ensures that each slice receives tailored optimization strategies depending on their type, maximizing their efficiency. While intra-CSP data sharing presents challenges, proposing inter-CSP slice data sharing introduces additional complexities. To enhance data security and privacy, the framework utilizes Blockchain as a distributed ledger, ensuring immutability, and trust among CSPs. The EDgE architecture is simulated using production-grade Network Slice Management (NSM) and Cloud Orchestrator on real-world 4G datasets and NSL-KDD intrusion detection datasets. Test results show that EDgE rapidly converges to high accuracy, ultimately achieving 97.6% across different scenarios, outperforming legacy systems and enhancing the security of the NSM ecosystem.

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results