Veerapu Goutham

Veerapu Goutham

Auto-generated profile, based on publicly available information

Veerapu Goutham

School of Electronics Engineering, Vellore Institute of Technology

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

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

Representative Publications

View All Publications
28 May 2024 in IEEE Access
Verified
Journal Article
Open access
KRKonkala Venkateswarlu Reddy
Konkala Venkateswarlu Reddy
BRB. S. Karthikeya Reddy
B. S. Karthikeya Reddy
VGVeerapu Goutham
Veerapu Goutham
MMMahesh Miriyala
Mahesh Miriyala
JNJ. S. Nisha
J. S. Nisha
VRVarsha Ramkumar
Varsha Ramkumar
The relentless global population growth and the ever-increasing food demand pose formidable challenges to the agricultural sector. Farmers grapple with the ongoing challenge of wildlife-induced crop damage and human loss, which not only impedes food production but also exacerbates supply and demand imbalances. However, the rise of TinyML enables Edge AI as a promising avenue for implementing resource-efficient deep learning techniques on low-end edge devices. In this paper, we introduce an innovative solution that harnesses the power of Edge AI using tinyML-based deep learning algorithms in conjunction with the Internet of Things (IoT) for animal intrusion detection and deterrence system. The proposed system is developed to create remotely managed defense system tailored to safeguard vast agricultural expanses. It integrates a laser detection system and an AI-CAM with light weight deep learning algorithms for animal intrusion detection and classification. This system also ensures efficient animal deterrence and real-time monitoring for farmers, enabling them to assess the situation with the assistance of an intelligent rover build using IoT. This work emphasizes on proposing a light-weight deep learning model named EvoNet for animal classification. Results reveal that the proposed model achieves the highest accuracy at 96.7%, surpassing other models presented in this paper. However, for edge devices where compact file sizes are crucial, the model also offers comparable accuracy with file sizes as low as 1.63MB with the help of pruning and quantization techniques. This conceptualized solution has the potential to revolutionize agricultural wildlife management, ushering in a new era of crop protection and economic resilience.

22 February 2021 in IEEE Access
Verified
Journal Article
Open access
VGVeerapu Goutham
Veerapu Goutham
HPHarigovindan V P
Harigovindan V P
In this paper, we propose non-orthogonal multiple access based cooperative relaying strategy (NOMA-CRS) for underwater acoustic sensor networks (UASNs). We analyse the performance of NOMA-CRS for both shallow and deep water scenarios, under imperfect channel state information (I-CSI) as well as imperfect successive interference cancellation (I-SIC). We derive mathematical expressions for ergodic rate, outage probability as well as the energy efficiency of NOMA-CRS in UASNs by considering the underwater specific characteristics, such as distance-dependent usable bandwidth, acoustic spreading, propagation loss, and fading effects. We compare the performance of NOMA-CRS with the widely used decode-and-forward based CRS in the UASNs. From the results, it is evident that NOMA-CRS can achieve significant improvement in ergodic sum rate and energy efficiency. But the outage performance is slightly degraded for the proposed scheme. Our results show that I-CSI and I-SIC have a significant impact on the performance of the NOMA-CRS. We also investigate the impact of relay position, wind speed as well as shipping activities on the performance of NOMA-CRS under the realistic underwater scenario. Results show that high-speed winds and high shipping activities severely degrade the performance of ergodic sum rate of the NOMA-CRS. Implementation of NOMA-CRS requires CSI at the transmitter. However, acquiring perfect CSI at the transmitter is a challenging task in time-varying multi-path underwater acoustic channels. As a solution, we also propose space-time block coded NOMA-CRS (STBC-NOMA-CRS) for UASNs, which can be implemented without CSI at the transmitter. Extensive simulation studies are conducted to corroborate the analytical findings.

7 February 2025 in IEEE Access
Verified
Journal Article
Open access
LALincy Annet Abraham
Lincy Annet Abraham
GPGopinath Palanisamy
Gopinath Palanisamy
VGVeerapu Goutham
Veerapu Goutham
The extremely low tumor prevalence in the general population makes accurate image-based tumor identification and classification a significant obstacle for medical researchers. Timely therapies and better patient outcomes depend on the capacity to identify cancers in anomalous circumstances. This research seeks to solve the problem by identifying and categorizing brain tumors in difficult cases using deep learning (DL) methods. Hence this paper, Dilated Convolution and YOLOv8 Feature Extraction Network (DC-YOLOv8FEN) is proposed to improve tumor detection accuracy. To begin, the feature extraction network (FEN) is improved with the help of the self shuffle attention (SSA), and the feature map is maximized with the help of vision transformer block. Secondly, a dual feature pyramid network (DFPN) is built to provide more discriminative data for dynamic sparse attention mechanism to extract features from the shallow network and top-down routes to direct the following network modules to fuse features. In addition, the DC-YOLOv8FEN uses a position regression loss function and adds Wise-IoU’s dynamic non-monotonic focus mechanism to make it even better. Based on the experimental results, the proposed method achieved precision of 95.2%, recall rate of 90.5%, accuracy of 99.5% and F1-score of 96.5% which evinces the outstanding performance compared with the existing methods.Radiologists may use this technology for better identification and diagnose brain tumors in real-time medical imaging systems, which improves patient outcomes and reduces diagnostic delays.

1 August 2025 in Scientific Reports
Verified
Journal Article
Open access
PDF file
LALincy Annet Abraham
Lincy Annet Abraham
GPGopinath Palanisamy
Gopinath Palanisamy
GVGoutham Veerapu
Goutham Veerapu
A crucial area of research in the field of medical imaging is that of brain tumor classification, which greatly aids diagnosis and facilitates treatment planning. This paper proposes DenseNet169-LIME-TumorNet, a model based on deep learning and an integrated combination of DenseNet169 with LIME to boost the performance of brain tumor classification and its interpretability. The model was trained and evaluated on the publicly available Brain Tumor MRI Dataset containing 2,870 images spanning three tumor types. Dense169-LIME-TumorNet achieves a classification accuracy of 98.78%, outperforming widely used architectures including Inception V3, ResNet50, MobileNet V2, EfficientNet variants, and other DenseNet configurations. The integration of LIME provides visual explanations that enhance transparency and reliability in clinical decision-making. Furthermore, the model demonstrates minimal computational overhead, enabling faster inference and deployment in resource-constrained clinical environments, thereby highlighting its practical utility for real-time diagnostic support. Work in the future should run towards creating generalization through the adoption of a multi-modal learning approach, hybrid deep learning development, and real-time application development for AI-assisted diagnosis.

16 December 2024 in IEEE Access
Verified
Journal Article
Open access
MGMegha Gupta
Megha Gupta
SGSneha Gupta
Sneha Gupta
GPGopinath Palanisamy
Gopinath Palanisamy
JNJ. S. Nisha
J. S. Nisha
VGVeerapu Goutham
Veerapu Goutham
GNGanesh R. Naik
Ganesh R. Naik
This study examines the application of color fundus imaging for detecting both ocular and non-ocular diseases using recent advancements in machine learning techniques. Fundus images provide essential diagnostic information for a range of retinal and systemic conditions. The review systematically explores various methods for diagnosing major eye diseases like diabetic retinopathy (DR), age related macular degeneration (AMD), glaucoma and retinal vein occlusion (RVO) as well as non-ocular disorders like cardiovascular disease (CVD) using fundus images. It includes a thorough comparison of machine learning and deep learning models, assessing their performance through metrics such as accuracy, sensitivity, specificity, precision, recall, and Area Under the Curve (AUC). The study also incorporates a dataset comparison table, evaluating the attributes and suitability of different datasets for specific diagnostic tasks. EfficientNetB0 is highlighted as a top performer for diabetic retinopathy detection, while methods like CNN-LSTM combinations, Grad-CAM, and Swin Transformers show promise in detecting AMD, glaucoma, and RVO, respectively. Additionally, models such as CNNs, DNNs, and DenseNet-169 have been effective for CVD detection through analysis of retinal images. The review examines strategies to advance detection techniques for each disease, with a particular emphasis on future research directions aimed at improving early disease detection.

2 June 2020 in Iranian Journal of Science and Technology, Transactions of Electrical Engineering
Verified
Journal Article
Closed access
PDF file
VGVeerapu Goutham
Veerapu Goutham
HPHarigovindan V P
Harigovindan V P