G. Sowmya

International Institute of Information Technology, Hyderabad

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

P. Thangavel

University of Madras

Ali Rohan

International Institute of Information Technology, Hyderabad

Rakshit Naidu

Oxford Dental College and Hospital

Varun Shankar

University of Utah

V. S. Shankar Sriram

SASTRA University

Manchalla O V P Kumar

B. Subbarayudu

Jawaharlal Nehru Technological University, Hyderabad

K. Venkata Subbareddy

Osmania University

N. Ramadevi

SRKR Engineering College

P Laxmikanth

Cooperative Institutions

International Institute of Information Technology, Hyderabad

Jawaharlal Nehru Technological University, Hyderabad

University of Madras

Anna University, Chennai

Osmania University

Electronics Corporation of India

SRM University

SRM Institute of Science and Technology

Visvesvaraya Technological University

Yes Technologies (United States)

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Residual Neural Networks for Human Action Recognition from RGB-D Videos

Journal of Image and Graphics (2023)

K. Venkata Subbareddy B. Pavani G. Sowmya N. Ramadevi

Recently, the RGB-D based Human Action Recognition (HAR) has gained significant research attention due to the provision of complimentary information by different data modalities. However, the current models have experienced still unsatisfactory results due to several problems including noises and view point variations between different actions. To sort out these problems, this paper proposes two new action descriptors namely Modified Depth Motion Map (MDMM) and Spherical Redundant Joint Descriptor (SRJD). MDMM eliminates the noises from depth maps and preserves only the action related information. Further SRJD ensures resilience against view point variations and reduces the misclassifications between different actions with similar view properties. Further, to maximize the recognition accuracy, standard deep learning algorithm called as Residual Neural Network (ResNet) is used to train the system through the features extracted from MDMM and SRJD. Simulation experiments prove that the multiple data modalities are better than single data modality. The proposed approach was tested on two public datasets namely NTURGB+D dataset and UTD-MHAD dataset. The testing results declare that the proposed approach is superior to the earlier HAR methods. On an average, the proposed system gained an accuracy of 90.0442% and 92.3850% at Cross-subject and Cross-view validations respectively.

RGB color model

10.18178/joig.11.4.343-352

Cite

Citations (0)

Design and Implementation of Efficient Block Cipher

Korkonda Srivaishnavi G. Sowmya P Laxmikanth Manchalla O V P Kumar Kiran Mannem

10.1109/iciteics61368.2024.10625542

Cite

Citations (0)

Convergence of a Finite-Time Zhang Neural Network for Moore–Penrose Matrix Inversion

Advances in intelligent systems and computing (2021)

G. Sowmya P. Thangavel

Zhàng

Matrix (chemical analysis)

Activation function

Inverse function

Sine

10.1007/978-981-15-8443-5_68

Cite

Citations (2)

A novel hybrid Zhang neural network model for time-varying matrix inversion

Engineering Science and Technology an International Journal (2021)

G. Sowmya P. Thangavel Varun Shankar

A new hybrid Zhang neural network (HZNN) model is formulated to solve time-varying matrix inversion problem. The network is designed such that it encompasses the goodness of both gradient neural network (GNN) and Zhang neural network (ZNN). The stability of the network and convergence results are established theoretically. The rate of convergence of HZNN model is much faster compared to the classical GNN model and ZNN model for time-varying matrix inversion problem. Nonlinear activation functions for HZNN need not have to be odd and monotonically increasing for the network to converge, however with certain restrictions. Simulations are carried out to test the efficiency of HZNN over GNN and ZNN. Also, HZNN with odd and non-odd activation functions are examined for convergence. It is observed that the network converges much faster with the proposed non-odd function 3 (nof3). Furthermore, the HZNN model with odd and non-odd functions are applied to the kinematic control of a two-link planar manipulator for tracking a circular path.

Activation function

Zhàng

Matrix (chemical analysis)

10.1016/j.jestch.2021.05.013

Cite

Citations (8)

Reversible data hiding in encrypted images based on side matching and enhanced complexity function

2017 International Conference on Algorithms, Methodology, Models and Applications in Emerging Technologies (ICAMMAET) (2017)

G. Sowmya P. Thangavel

Cloud computing has developed to be the most promising business model in recent times owing to access business applications or infrastructure resources at a negligible cost but the issue of security still lies as a question to be answered since storing and processing of data are done by a third-party data centers. Reversible data hiding in images guaranties security to data accumulated at a third-party center. Zhang proposed a complexity function that neglects borders of the image block. Hong et al. worked on the side matching technique and evaluated complexity function considering two adjacent pixels of the image block. Liao et al. organize image block of three different groups containing two, three and four neighborhoods for the computation of complexity function. In our proposed method, instead of utilizing the whole image block for data embedding, selection of pixels are based on the unencrypted fourth least significant bit (LSB) layer because it is nearly random. Further, a new complexity function is proposed considering every pixels of the image. The proposed method decreases the error rate during bit extraction when compared with existing approaches.

Block size

10.1109/icammaet.2017.8186730

Cite

Citations (0)

The video hiding technique based on DWT and genetic algorithm with boolean functions

G. Sowmya K. Meghana B. Subbarayudu Rakshit Naidu Keerthi Gurushanth

Steganography is a method of hiding a secret message within an ordinary message, where as the goal of cryptography is to make data unreadable by an unauthorized persons. Video steganography is expansion of image steganography where video files can simply viewed in a sequence of images. There are many difference between video and image steganography. As the video content is very dynamic and have less chances of detecting the hiding data compare to image steganography. So it is helpful than other existing stegnographic techniques because of it high security. It can also hold large amount of secret data. We introduced a technique called Discrete Wavelet Transformation (DWT) and Genetic algorithm to convert the data into a video file to make the video Steganography, which is an efficient method of embedding information into a carrier file. The DWT technique convert the cover image as 4 × 4 level of blocks in which each levels LL, LH, HL, and HH sub band images are developed. The current paper uses the DWT technique where the data is hidden in specific location of selected structure. Genetic algorithm is also introduced to minimize the selected frame and improve the security so that probability of detecting the hidden information into cover media is difficult. Boolean functions have been generated from different position arrays and measures the quality of the video indifferent byte streams by stego video.

Steganography tools

10.1109/ngct.2016.7877528

Cite

Citations (2)

Decoding Sentiments: A Comprehensive Exploration of Real-Time Emotion Recognition from Text Through Deep Learning

2022 8th International Conference on Advanced Computing and Communication Systems (ICACCS) (2024)

G. Sowmya R. Akhilesh Ali Rohan Garikapati Kausthub Rao Anurag Nayak

10.1109/icaccs60874.2024.10717285

Cite

Citations (0)

GRADIENT COMPASS-BASED ADAPTIVE MULTIMODAL MEDICAL IMAGE FUSION

Zenodo (CERN European Organization for Nuclear Research) (2022)

G. Sowmya

Compass

10.5281/zenodo.7985220

Cite

Citations (0)

A novel finite-time complex-valued zeoring neural network for solving time-varying complex-valued Sylvester equation

Journal of the Franklin Institute (2022)

G. Sowmya V. S. Shankar Sriram P. Thangavel

Activation function

Sign function

Error function

10.1016/j.jfranklin.2022.10.022

Cite

Citations (0)

Noise Removal in Image Using LPG-PCA (Local Pixel GroupingPrinciple Component Analysis) Algorithm

IOSR Journal of Electronics and Communication Engineering (2014)

Gadhiraju Hari Priya Bala Venkatesh A. Srivani G. Sowmya N. Rajasekhar

In this paper an effective algorithm for noise removal in an image is obtained by using PCA (principal component analysis) with LPG (Local Pixel Grouping).This technique ensures the preservation of image local structure.Here the pixels and its neighbors are treated as vector variables whose training samples are selected from local windows using block matching based LPG.This ensures only the similar samples are selected for the PCA transformation so that the desired local characteristics are only preserved with considerable noise reduction.The LPG -PCA algorithm is performed twice to enhance the quality of an image.The first iteration would remove the noise considerably and the second iteration would preserve the image features like edges etc.The LPG-PCA algorithm will adaptively adjust the noise level of an image unlike WT (Wavelet Transformation).Several experimental results show the effectiveness of the proposed algorithm.

Component (thermodynamics)

10.9790/2834-09153642

Cite

Citations (1)