Welcome


I am a Doctoral student at Machine Vision Lab, Department of Computer Science & Engineering, Indian Institute of Technology Roorkee, INDIA. Prof. Balasubramanian Raman is my advisor.  

My research lies at the intersection of Multimedia Processing, Cloud Computing and Security. My recent works have focused on privacy-preserving multimedia computing, especially colorization of digital images/videos over cloud. I am interested in learning how digital multimedia can be processed in a secure manner without revealing any useful information at third party cloud servers. Any discussion regarding research in Privacy-Preserving Data Processing, Cloud Computing and Security is most welcome. 


Research Topics

Privacy-Preserving Multimedia Computing

Cloud computing provides cost effective pay-as-you-go services with business continuity. Users are relieved from infrastructure maintenance responsibility and can accomplish the services directly offered by cloud service providers at reasonable cost. As a result, new cloud enabled gadgets are evolving nowadays. Regardless of these benefits, users are always concerned when moving their personal data to the cloud. Hence, new techniques are required to process user's data without revealing any information over third party cloud server and falls in the area of Privacy Preserving Computing/Encrypted Domain Processing.


Example based privacy-preserving image color transfer

Considering the importance of privacy-preserving multimedia computing, we have developed novel methods to process multimedia colors in the encrypted domain. The proposed methods consist of processing image/video colors in the encrypted domain in a manner such that the same effects can be perceived as of processing over plain text data. Initially, an example image needs to be selected as the reference and colors of the underlying test image are then transformed as per colors of the reference image. The proposed work finds vast scope in various applications where image color plays a vital role to enhance the performance of the underlying algorithm (e.g. panoramic and 360-degree formulation, machine vision systems, medical imaging etc.).




Privacy-preserving smart surveillance using local color correction

Due to rapid growth of cloud based surveillance systems, we found that image color enhancement is a much needed necessity. Nowadays, surveillance consists of smart cameras connected to cloud servers leveraging full-fledged processing and streaming of captured scene to the authorized person remotely. The images captured from camera sensor are processed at their best to cope with the human/machine vision system. However, due to inconsistent lightning conditions, it becomes a necessary task to post-process the captured multimedia for better appearance. Moreover, surveillance data is of very large size and suffers from the problem of local storage. In this situation, massive storage pools available at cloud server are one of the best options to adopt. Although, since cloud servers are developed and maintained by third parties, there is a risk of privacy for highly confidential data like surveillance images/videos. Addressing these problems, we have worked on a novel cloud based privacy-preserving approach for image color enhancement. Advancing the seminal work where colors of the underlying image are transformed as per the colors of a reference image, we propose to enhance image colors using its local color characteristics in a privacy-preserving manner over the cloud. The proposed approach finds vast scope in privacy-preserving cloud based smart surveillance, where colors of the captured multimedia are enhanced in an automatic manner without revealing any useful information over third party cloud servers.

               

                

Research Publications
  1. Amitesh Singh Rajput and Balasubramanian Raman, Privacy-preserving smart surveillance using local color correction and optimized ElGamal cryptosystem over cloud, accepted for publication in IEEE International Conference on Cloud Computing (IEEE CLOUD), affiliated with IEEE World Congress on Services (IEEE SERVICES), to be organized at Milan, ITALY during July 2019 (Acceptance rate  20.8%).
  2. Amitesh Singh Rajput and Balasubramanian Raman, Example based privacy-preserving video color grading, accepted for publication in Handbook of Multimedia Information Security: Techniques and Applications, Multimedia Systems and Applications, Springer, 2019.
  3. Amitesh Singh Rajput and Balasubramanian Raman, Cloud based image color transfer and storage in encrypted domain, Multimedia Tools and Applications, Springer, Vol. 77, No. 16, pp. 21509-21537, 2018.
  4. Vishesh Kumar Tanwar, Amitesh Singh Rajput, Balasubramanian Raman and Rama Bhargava, Privacy preserving image scaling using 2D bicubic interpolation over the cloud, IEEE International Conference on Systems, Man and Cybernetics (IEEE SMC), 2018, Miyazaki, Japan.
  5. Amitesh Singh Rajput and Balasubramanian Raman, CryptoCT: Towards privacy preserving color transfer and storage over cloud, Multimedia Tools and Applications, Springer, Vol. 77, No. 18, pp. 24223-24245, 2018.
  6. Amitesh Singh Rajput and Balasubramanian Raman, Color me, store me, know me not: Supporting image color transfer and storage in encrypted domain over cloud, In Proceedings of IEEE International Conference on Multimedia & Expo Workshops (ICMEW), pp. 291-296, July 2017, Harbour Grand, HONG KONG.
  7. Vikrant Kumar Saini, Amitesh Singh Rajput and Balasubramanian Raman, Robust image sharing scheme using one dimensional chaotic logistic map, in Proceedings of Second International Conference on Computer Vision & Image Processing (CVIP), pp. 157-168, September 2017, GNEC IIT Roorkee, INDIA.
  8. Arun Singh Pundir, Himanshu Buckchash, Amitesh Singh Rajput, Vishesh Kumar Tanwar and Balasubramanian Raman, Fire Detection using Dense Trajectories, in Proceedings of Second International Conference on Computer Vision & Image Processing (CVIP), pp. 211-221, September 2017, GNEC IIT Roorkee, INDIA.
  9. Amitesh Singh Rajput, Priyanka Singh and Balasubramanian Raman, CryptFine: Towards secure cloud based filtering in encrypted domain, In Proceedings of ISEA IEEE Asia Security and Privacy Conference (ISEASP), pp. 1-8, January 2017, NIT Surat, Gujarat, INDIA.
  10. Amitesh Singh Rajput and Mansi Sharma, A Novel Image Encryption and Authentication Scheme Using Chaotic Maps, in Proceedings of Advances in Intelligent Informatics, Springer, organized with IEEE International Conference on Advances in Computing, Communications and Informatics (ICACCI), pp. 277-286, September 2015, GCET, Greater Noida, INDIA.
  11. Amitesh Singh Rajput, Nishchol Mishra and Sanjeev Sharma, Towards the growth of image encryption and authentication schemes, in Proceedings of IEEE International Conference on Advances in Computing, Communications and Informatics (ICACCI), pp. 454-459, September 2013, SJCE, Mysore, INDIA.
Awards and Scholarships
  1. Registration support from Microsoft Research to present my research paper at the prestigious International Conference on Multimedia and Expo-Workshop on Multimedia and the Cloud, organized at Harbour Grand, HONG KONG, 2017.
  2. Travel support from Ministry of Electronics and Information Technology (MeitY), Government of INDIA to attend the prestigious International Conference on Multimedia and Expo, organized at Harbour Grand, HONG KONG, 2017.
  3. Visvesvaraya PhD scholarship provided by Ministry of Electronics and Information Technology (MeitY), Government of INDIA, (December 2015 - till date).
  4. Qualified EMC Academic Associate certification in Cloud Infrastructure & Services, organized by EMC India, 2014.
  5. MHRD-GATE Scholarship for Master of Technology course, (2011 - 2013).
  6. Worked with Jr. Scienti.c officer, Madhya Pradesh Council of Science and Technology (MPCST), for developing a website for library management of MPCST, Bhopal, INDIA.

Synthetic image set for single dominating color tone

To access the dataset, please send an e-mail to arajput@cs.iitr.ac.in.