Thounaojam Rupachandra Singh
@manipuruniv.ac.in
Assistant Professor, Computer Science Department
Manipur University
EDUCATION
MCA, NET, PhD
RESEARCH INTERESTS
Network, SDN, Image processing
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Thiyam Romita Chanu, Th. Rupachandra Singh, and Kh. Manglem Singh
Springer Science and Business Media LLC
Thounaojam Rupachandra Singh and Khaidem Bikramjit Meitei
AIP Publishing
Oinam James, Th. Rupachandra Singh, and T.Romen Singh
IEEE
Image High Resolution takes a principal function to the region of the photograph filtering to enhance the photograph and video frames for Enhanced Resolution in gadgets such as laptops and smartphones. This presents a deep learning technique for enhancing resolution for single picture super resolution. The method directly learns an end to end mapping between low quality and high-quality resolution pictures. A deep convolutional neural network represents the mapping, which takes the poor resolution picture as input and produces a high quality resolution image.
Irengbam Tilokchan Singh, Thounaojam Rupachandra Singh, and Tejmani Sinam
IEEE
Round Robin Algorithm plays a crucial role in load balancing of server farms. This paper’s primary focus is to benchmark the Round Robin algorithm in Server Load Balancing (SLB) using Software Defined Networking (SDN). Experimental results show that the Round Robin algorithm gives 100% availability of the servers. Load balancing is done among the live servers from the server pool. In our experiment, the Round Robin load-balancing algorithm is implemented using a POX controller and an OpenFlow switch.
Manganleima Moirangthem, Thounaojam Rupachandra Singh, and Th. Tangkeshwar Singh
IEEE
To test the ideas of machine learning, effects of preprocessing and Region of Interest (ROI) segmentation on the performance of Content Based Medical Image Retrieval (CBMIR) Systems, we have developed a CBMIR prototype using the MATLAB environment. Brain MRI image dataset from Kaggle have been utilized to train and test the Resnet50 Convolutional Neural Network (CNN) which are fed with data, first without preprocessing and without the concepts of ROI and then with data which were preprocessed and ROI segmented. The two different approaches have been compared by measures of accuracy, precision and recall of the classification thus achieved.
Moirangthem Marjit Singh, Nishigandha Dutta, Thounaojam Rupachandra Singh, and Utpal Nandi
Springer Singapore
Wormhole attack is a harmful attack that disrupts the normal functioning of the network by manipulating routing protocols and exhausting network resources. The paper presents a technique that detects the presence of wormhole attack in wireless sensor network (WSN) using artificial neural network (ANN). The proposed technique uses connectivity information between any two sensor nodes as the detection feature. The proposed technique has been implemented considering the deployment of sensor nodes in the wireless sensor network area under uniform, Poisson, Gaussian, exponential, gamma & beta probability distributions. The proposed technique does not require any additional hardware resources and gives a comparatively high percentage of detection accuracy.
Khaidem Bikramjit Meitei and Thounaojam Rupachandra Singh
Institute of Advanced Scientific Research
Th. Rupachandra Singh, Kh. Manglem Singh, and Sudipta Roy
Elsevier BV
Th. Rupachandra Singh, Kh. Manglem Singh, and Sudipta Roy
IEEE
A novel video watermarking scheme based on visual cryptography and scene change detection in discrete wavelet transform domain is proposed. We start with a complete survey of the current image and video watermarking technologies, and have noticed that majority of the existing schemes are not capable of resisting all attacks. We propose the idea to use different parts of a single watermark into different scenes of a video for generation of the owner's share from the original video based on the frame mean in same scene and the binary watermark, and generation of the identification share based on the frame mean of probably attacked video. These two shares after stacking can reveal the copyright ownership. Experiments are conducted to verify the robustness through a series of experiments.