Lakshmi Hassan Nagaraja

Lakshmi Hassan Nagaraja

Auto-generated profile, based on publicly available information

Lakshmi Hassan Nagaraja

CVR College of Engineering

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 April 2025 in Journal of Machine and Computing
Verified
Journal Article
Open access
SHShaik Jaffar Hussain
Shaik Jaffar Hussain
RBRupa Devi B
Rupa Devi B
AEAnant Shankar E
Anant Shankar E
LNLakshmi H N
Lakshmi H N
RKRangaswamy K
Rangaswamy K
SKSuneelgoutham Karudumpa
Suneelgoutham Karudumpa
Advanced diagnostic tools are essential for aerospace transportation systems and automotive industries and industrial manufacturing facilities since operational efficiency requirements and safety needs demand failure prediction tools. Systems that use traditional diagnostic methods depend on centralized architectures that show limitations regarding scalability while being unable to overcome subsystem failure events effectively. The research presents Gossip Neural Network (GNN) as a decentralized deep learning (DL) system which determines Remaining Useful Life (RUL) duration in distributed mechanical engine systems. The GNN combines Convolutional Neural Networks (CNNs) and Long short- term Memory (LSTM) network layers to identify short-term sensor anomalies in addition to capturing long-term sensor degeneration patterns in sensor data. A gossip-based protocol allows the GNN to facilitate distributed engine subsystems which train a shared model together through peer-to-peer collaborations without needing central control. The assessment of the proposed framework using CMAPSS data proves its exceptional capability for RUL prediction alongside reliable accuracy and low error rates. The GNN demonstrated excellence in different datasets through R² results between 92.43% and 94.57% and RMSE results within 12.77 to 12.87 which demonstrates its effectiveness in handling realistic operational environments. The GNN provides an encouraging solution for time-sensitive fault detection in distributed systems which facilitates efficient predictive maintenance across large engineering applications.

1 October 2024 in International Journal of Electrical and Computer Engineering (IJECE)
Verified
Journal Article
Open access
PDF file
HLH. N. Lakshmi
H. N. Lakshmi
TRThaduri Venkata Ramana
Thaduri Venkata Ramana
LKLnc Prakash K
Lnc Prakash K
LRL. Kiran Kumar Reddy
L. Kiran Kumar Reddy
KRKachapuram Basava Raju
Kachapuram Basava Raju
Ensemble learning stands out as a widely embraced technique in machine learning. This research explores the application of ensemble learning, including ensemble clustering, to enhance the precision of cluster analysis for datasets with multiple attributes and unclear correlations. Employing a majority voting-based ensemble clustering approach, specific techniques such as k-means clustering, affinity propagation, mean shift, BIRCH clustering, and others are applied to defined datasets, leading to improved clustering results. The study involves a comprehensive comparative analysis, contrasting ensemble clustering outcomes with those of individual techniques. The process of improving cluster identification accuracy encompasses data collection, pre-processing to exclude irrelevant elements, and the application of standard clustering algorithms. The task includes defining the optimal number of groups before comparing clustering models. Additionally, a combined model is constructed by merging BIRCH clustering and mean shift clustering, leveraging their advantages to enhance overall clustering strength and accuracy. This research contributes to advancing ensemble learning and ensemble clustering methodologies, offering improved accuracy, and uncovering hidden patterns in complex datasets.

25 April 2025
Verified
Book Chapter
Closed access
BDB. Rupa Devi
B. Rupa Devi
LNLakshmi H N
Lakshmi H N
MTMary Teresa
Mary Teresa
GSG. Suseela
G. Suseela
PBPenubaka Balaji
Penubaka Balaji
Wearable technology and predictive algorithms are transforming neurorehabilitation by offering data-driven, personalized solutions for patient recovery. With the growing prevalence of neurological impairments, this study investigates how real-time sensor data from wearable devices can drive adaptive rehabilitation programs tailored to individual needs. These technologies enable immediate feedback, optimizing therapy based on patient performance to enhance recovery outcomes. The research highlights the efficacy of wearable devices and predictive models in improving rehabilitation strategies through data analytics, AI, and real-world case studies. Challenges such as sensor accuracy, algorithm refinement, and system integration are addressed, emphasizing the need for future advancements to create efficient and effective neurorehabilitation processes.

5 July 2025 in Journal of Machine and Computing
Verified
Journal Article
Open access
VYVishnu Tej Y
Vishnu Tej Y
AAAshwitha A
Ashwitha A
LNLakshmi H N
Lakshmi H N
VBVuppala Balaji
Vuppala Balaji
SGSuryanarayana G
Suryanarayana G
SMSirish Kumar M
Sirish Kumar M
A key element of sustainable development is efficient trash classification, which aims to minimize environmental damage and expedite recycling procedures. In addition to being time-consuming, traditional human sorting methods are prone to mistakes, which makes waste management systems less effective. Automated garbage classification has attracted so much attention as AI, especially ML and DL, has grown. However, because they frequently rely on small-scale datasets and traditional architectures, many of the models that are now in use have issues with generalization, poor performance, and high error rates. This work presents a hybrid deep learning system that combines an autoencoder with a vision transformer (ViT) to address these issues. By efficiently capturing local and global data, our design improves classification robustness and accuracy across various waste types. Our model was trained and assessed using a sizable and varied dataset to enhance generalization to real-world scenarios. According to experimental data, the suggested model achieves a precision of 96.72%, a recall of 96.21%, an F1-score of 96.46%, and a balanced accuracy of 96.48%, outperforming some cutting-edge CNN-based architectures. Furthermore, sophisticated measures like Cohen's Kappa (95.90%) and Matthews Correlation Coefficient (MCC = 94.91%) confirm the dependability of our solution. Lastly, by successfully implementing the model in an inference pipeline, we show that it is ready for real-world deployment and that it has the potential to promote sustainable development goals through scalable, intelligent waste management.

5 January 2024 in Journal of Machine and Computing
Verified
Journal Article
Open access
GSGunji Sreenivasulu
Gunji Sreenivasulu
LNLakshmi H N
Lakshmi H N
MTMuni Kumari T
Muni Kumari T
APAnjaiah P
Anjaiah P
SASuresh A
Suresh A
AJAvanija J
Avanija J
The timely and precise identification of traffic signs is essential for maintaining the effectiveness and safety of contemporary roads, particularly in light of the increasing number of self-driving cars. Conventional image processing methods have faced challenges because to the intricate and fluctuating variables present in real-world settings, including various signage, erratic weather, and inconsistent illumination. This study utilizes recent breakthroughs in deep learning, particularly the YOLOv8 (You Only Look Once version 8) model, to tackle these difficulties. YOLOv8 incorporates cutting-edge neural network architectural advancements, such as an anchor-free detection methodology, adaptive spatial feature pooling, and dynamic neural configurations. In order to further increase detection efficiency and accuracy, this study presents two innovative models, YOLOv8-DH and YOLOv8-TDHSA. These models make use of improvements such decoupled heads and transformer-based self-attention mechanisms. Experimental results indicate that the suggested models substantially surpass current deep learning models, attaining enhanced performance across multiple measures, including accuracy, recall, F-score, and mean average precision (mAP). This research enhances traffic sign detecting technology, facilitating the development of safer and more intelligent transportation systems.

21 May 2025 in medRxiv
Verified
Preprint
Open access
PDF file
SASyed Thouheed Ahmed
Syed Thouheed Ahmed
SYSatheesha Tumakur Yoga
Satheesha Tumakur Yoga
LNLakshmi Hassan Nagaraja
Lakshmi Hassan Nagaraja
SMSandeep kumar M
Sandeep kumar M
RSR. Sangeetha
R. Sangeetha
SMSaurav Mallik
Saurav Mallik
Lung cancer detection and treatment is processed using upgraded medical tools and radiology experts from multiple medical data sources. The challenge in decision making is dependent on experts understanding on a given electronic health records or datasets. In this paper, we have proposed an improvised approach of lung cancer classification based on intensity driven RoI selection from the Lung Images Database Consortium Image Collection (LIDC-IDRI), Cancer Imaging Archive (CIA) datasets. The technique is developed on label customization and annotating the vulnerable RoI regions. The approach is optimized for higher dimensionality mapping of RoIs. The technique deploys a feedback-based upgrading and monitoring approach via transfer learning framework. The trained dataset from RoI optimizer is updated to customized learning models for decision transfer and decision-making capabilities. The proposed technique is deployed on CoVNET framework and has demonstrated the accuracy of 97.84% under a 60:40 training testing-based learning model.