Quantum Computing in Cybersecurity

16Image Filtering Based on VQA with Quantum Security

Avni Burman*, Bhushan Bawaskar, Harsh Dindorkar and Hrithik Surjan

Department of Computer Science Engineering, Shri Vaishnav Vidyapeeth Vishwavidyalaya, Indore, India

Abstract

With the increasing application of VQA (Visual Question Answering), a wide range of applications has been introduced and there are also several ongoing researches on VQA. In VQA we build and train our system to give the correct answer for the provided dataset, i.e., the image and question. We have to provide an image and question to the model but during the direct fusion of feature elements of both question (Qi) and image (Vi) there are several problems such as noise in the image and question. If the given image is noisy then there is no accuracy [3]. Both the elements (Qi, Vi) are not in the same space. To solve this we introduce Image filtering, Differential Network and its Fusion. In today’s world data is everything, and its protection should be our priority. So, to protect VQA, we are introducing RSA algorithm. RSA algorithm will ensure the security of VQA from any intrusion and will maintain its authenticity.

Keywords: Visual Question Answering (VQA), Differential Network (DN), image filtering, median filter, rsa algorithm, cryptography, quantum security

16.1 Introduction

VQA [1, 2] has flourished in recent years due to its numerous uses. The model effectively merges feature components from both the photos and the questions to complete the VQA ...