English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This is Btech thesis report on detection and purification of adverseriallyattacked images. A deep learning model is trained on certain training examplesfor various tasks such as classification, regression etc. By training, weightsare adjusted such that the model performs the task well not only on trainingexamples judged by a certain metric but has an excellent ability to generalizeon other unseen examples as well which are typically called the test data.Despite the huge success of machine learning models on a wide range of tasks,security has received a lot less attention along the years. Robustness alongvarious potential cyber attacks also should be a metric for the accuracy of themachine learning models. These cyber attacks can potentially lead to a varietyof negative impacts in the real world sensitive applications for which machinelearning is used such as medical and transportation systems. Hence, it is anecessity to secure the system from such attacks. Int this report, I focus on aclass of these cyber attacks called the adversarial attacks in which theoriginal input sample is modified by small perturbations such that they stilllook visually the same to human beings but the machine learning models arefooled by such inputs. In this report I discuss 2 novel ways to counter theadversarial attack using AutoEncoders, 1) by detecting the presence ofadversaries and 2) purifying these adversaries to make target classificationmodels robust against such attacks.

Btech thesis report on adversarial attack detection and purification of adverserially attacked images