Siddharth Ravindran
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Siddharth Ravindran has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Aghila Gnanasekaran
Department of CSE, National Institute of Technology Puducherry, Karaikal 609609, India
3 shared publications G. Aghila
National Institute of Technology Puducherry
2 shared publicationsAghila G.
National Institute of Technology Puducherry, India
1 shared publicationG Aghila
National Institute of Technology Puducherry
1 shared publication- Hemachandiran Shanmugam
Department of CSE, National Institute of Technology Puducherry, Karaikal 609609, India
1 shared publication S. Hemachandiran
National Institute of Technology Puducherry
1 shared publication
Latest Publications
- 17 March 2023
- Published by Springer Nature in International Journal of Information Technology
- Vol. 15 (3) , 1443-1452
A digital image colorimetry approach for identifying fuel types in downstream petroleum sector
- 17 March 2023
- Published by Springer Nature in International Journal of Information Technology
- Vol. 15 (3) , 1443-1452
A digital image colorimetry approach for identifying fuel types in downstream petroleum sector
- 24 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Computing
- Vol. 7 (4) , 825-838
A Fog-Assisted Framework for Intelligent Video Preprocessing in Cloud-Based Video Surveillance as a Service
Video preprocessing is one of the vital steps before extracting useful information from surveillance videos. This article identifies the key issue in cloud-based video surveillance as a service (VSaaS). If the recording environment is almost static and storing 24/7 video recording leads to additional storage, computational and network complexity to the cloud user. To address this problem, the existing architecture of the VSaaS is thoroughly studied and proposed a fog-assisted framework for preprocessing the incoming video. The two important functionalities of this framework are intelligent frame-selection and frame-reconstruction. During frame-selection, the lightweight background subtraction algorithm based on random projection with structural similarity index manages to effectively select the necessary frames to store in the cloud. This framework also provides an option for the user to successfully reconstruct the full-length video using frame inbetweening method. The proposed algorithm is tested with CDNET-2014 and LASIESTA data. The results are evaluated in-terms of running time, selected frames per video, percentage of reduction and quality of video experiencing. The result shows that the proposed frame-selection manages to reduce nearly 60% of storage requirement for the CDNET-2014 data. The frame reconstruction algorithm reconstructs the full-length video without affecting the quality of video experiencing.
- 24 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Computing
- Vol. 7 (4) , 825-838
A Fog-Assisted Framework for Intelligent Video Preprocessing in Cloud-Based Video Surveillance as a Service
Video preprocessing is one of the vital steps before extracting useful information from surveillance videos. This article identifies the key issue in cloud-based video surveillance as a service (VSaaS). If the recording environment is almost static and storing 24/7 video recording leads to additional storage, computational and network complexity to the cloud user. To address this problem, the existing architecture of the VSaaS is thoroughly studied and proposed a fog-assisted framework for preprocessing the incoming video. The two important functionalities of this framework are intelligent frame-selection and frame-reconstruction. During frame-selection, the lightweight background subtraction algorithm based on random projection with structural similarity index manages to effectively select the necessary frames to store in the cloud. This framework also provides an option for the user to successfully reconstruct the full-length video using frame inbetweening method. The proposed algorithm is tested with CDNET-2014 and LASIESTA data. The results are evaluated in-terms of running time, selected frames per video, percentage of reduction and quality of video experiencing. The result shows that the proposed frame-selection manages to reduce nearly 60% of storage requirement for the CDNET-2014 data. The frame reconstruction algorithm reconstructs the full-length video without affecting the quality of video experiencing.
- 12 October 2021
- Published by Springer Nature
- p. 415-423
Automation to Find Adulteration in Downstream Petroleum Monitoring Using Machine Learning: An Overview
- 12 October 2021
- Published by Springer Nature
- p. 415-423
Automation to Find Adulteration in Downstream Petroleum Monitoring Using Machine Learning: An Overview
- 18 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Letters of the Computer Society
- Vol. 3 (1) , 17-20
A Light Weight Background Subtraction Algorithm for Motion Detection in Fog Computing
A novel light weight background subtraction algorithm proposed in this letter addresses the issue of executing real time motion detection in fog computing, which is a resource constrained environment. The proposed algorithm compresses the original high dimensional video frame into low dimensional space using random projection(RP) and the motion detection is identified using the structural similarity (SSIM) index. The objective of the proposed algorithm is to select the frames with motion and discard the remaining frame in order to reduce the storage and computational cost for the cloud users. The proposed algorithm is evaluated using CDNET 2014 video data and the results are verified in terms of selected frames per video(SFPV), percentage of reduction(POR) and running time. From the experiment, it is evident that the proposed algorithm reduces nearly 67% of storage cost with trifling computational expense.
- 18 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Letters of the Computer Society
- Vol. 3 (1) , 17-20
A Light Weight Background Subtraction Algorithm for Motion Detection in Fog Computing
A novel light weight background subtraction algorithm proposed in this letter addresses the issue of executing real time motion detection in fog computing, which is a resource constrained environment. The proposed algorithm compresses the original high dimensional video frame into low dimensional space using random projection(RP) and the motion detection is identified using the structural similarity (SSIM) index. The objective of the proposed algorithm is to select the frames with motion and discard the remaining frame in order to reduce the storage and computational cost for the cloud users. The proposed algorithm is evaluated using CDNET 2014 video data and the results are verified in terms of selected frames per video(SFPV), percentage of reduction(POR) and running time. From the experiment, it is evident that the proposed algorithm reduces nearly 67% of storage cost with trifling computational expense.
- 6 February 2019
- Published by Springer Nature in National Academy Science Letters
- Vol. 43 (1) , 13-21
A Data-Independent Reusable Projection (DIRP) Technique for Dimension Reduction in Big Data Classification Using k-Nearest Neighbor (k-NN)
- 6 February 2019
- Published by Springer Nature in National Academy Science Letters
- Vol. 43 (1) , 13-21
A Data-Independent Reusable Projection (DIRP) Technique for Dimension Reduction in Big Data Classification Using k-Nearest Neighbor (k-NN)
- 1 January 2019
- Published by IGI Global Scientific Publishing
- p. 151-169
A Privacy-Preserving Feature Extraction Method for Big Data Analytics Based on Data-Independent Reusable Projection
Big data analytics is one of the key research areas ever since the advancement of internet technologies, social media, mobile networks, and internet of things (IoT). The volume of big data creates a major challenge to the data scientist while interpreting the information from raw data. The privacy of user data is an important issue faced by the users who utilize the computing resources from third party (i.e., cloud environment). This chapter proposed a data independent reusable projection (DIRP) technique for reducing the dimension of the original high dimensional data and also preserves the privacy of the data in analysis phase. The proposed method projects the high dimensional input data into the random low dimensional space. The data independent and distance preserving property helps the proposed method to reduce the computational complexity of the machine learning algorithm. The randomness of data masks the original input data which helps to solve the privacy issue during data analysis. The proposed algorithm has been tested with the MNIST hand written digit recognition dataset.
- 1 January 2019
- Published by IGI Global Scientific Publishing
- p. 151-169
A Privacy-Preserving Feature Extraction Method for Big Data Analytics Based on Data-Independent Reusable Projection
Big data analytics is one of the key research areas ever since the advancement of internet technologies, social media, mobile networks, and internet of things (IoT). The volume of big data creates a major challenge to the data scientist while interpreting the information from raw data. The privacy of user data is an important issue faced by the users who utilize the computing resources from third party (i.e., cloud environment). This chapter proposed a data independent reusable projection (DIRP) technique for reducing the dimension of the original high dimensional data and also preserves the privacy of the data in analysis phase. The proposed method projects the high dimensional input data into the random low dimensional space. The data independent and distance preserving property helps the proposed method to reduce the computational complexity of the machine learning algorithm. The randomness of data masks the original input data which helps to solve the privacy issue during data analysis. The proposed algorithm has been tested with the MNIST hand written digit recognition dataset.
- 1 April 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
A cost-effective approach towards storage and Privacy preserving for Intermediate data sets in Cloud Environment
Cloud computing offers pay-as-you-go model, where users only pay for their resource consumption. Many large applications utilize cloud computing. These applications generate a lot of essential intermediate results for future purpose. Storing all intermediate results is not a cost efficient approach. At the same time adversary may refer multiple intermediate result to steal the information. Likewise encrypting every part of intermediate results will increase computation cost for the user. The main aim of the system is to provide a cost effective approach for storing and providing privacy for the intermediate results.
- 1 April 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
A cost-effective approach towards storage and Privacy preserving for Intermediate data sets in Cloud Environment
Cloud computing offers pay-as-you-go model, where users only pay for their resource consumption. Many large applications utilize cloud computing. These applications generate a lot of essential intermediate results for future purpose. Storing all intermediate results is not a cost efficient approach. At the same time adversary may refer multiple intermediate result to steal the information. Likewise encrypting every part of intermediate results will increase computation cost for the user. The main aim of the system is to provide a cost effective approach for storing and providing privacy for the intermediate results.