Evaluating Critical Level Of Perturbations Required To Achieve Certain Fail Rate

The fundamental point of the current research is to measure the precise amount of perturbations (noise) needed to be added to a certain image in order for it to be misclassified by the network. The estimation would be achieved via the method of adversarial attack.

The outcome of the research is important, for they would be crucial while using
networks involving matters of great importance where incorrect results of the process would lead to detrimental consequences.