Dr. Amitesh Singh Rajput

amiteshrajput [at] nliu [dot] ac [dot] in

arajput [at] cs [dot] iitr [dot] ac [dot] in

Cyber Security Researcher, Machine Learning Enthusiast, and a Professor!

Research

Privacy-Preserving Computing

Data Computing! When we see this word, we start thinking about a machine with some data. Yes, it's definitely true. But, considering the wide spectrum of data computing and its various mechanisms, one cannot simply ignore the associated security and privacy risks. Cloud computing, distributed data analysis, healthcare, Video-Surveillance-as-a-Service (VSaaS), file transfer, IoT, federated learning, etc. involves huge security and privacy risks. Also, these facilities are significantly affected, involving huge customer concerns, if proper security and privacy mechanisms are not made. My research involves identifying problems related to secure data computing and devising solutions to them. The term "Privacy-Preserving" has a special meaning in this context.

Privacy-Preserving Computing involves processing the underlying data without revealing any sensitive information. This becomes very important when the data needs to be processed over some third-party servers, e.g., the cloud. The seamless on-demand remote services provided by cloud servers are really good to cherish. However, the massive data collection required for deep learning presents privacy issues. Devising privacy-preserving methods to deal with this issue is my top interest.

A few results

Privacy-Preserving Human Action Recognition using Deep CNN

Using image obfuscation for human action recognition using deep Convolutional Neural Network (CNN).

Privacy-Preserving Smart Surveillance using Local Color Correction

Improving image colors using its local features.

Secured biomedical model for effective skin lesion segmentation

Effective and secured segmentation of the lesion area.

Google Sites

Report abuse