Exploring Secure Machine Learning Through Payload Injection and FGSM Attacks on ResNet-50

TL;DR

This study examines the security vulnerabilities of ResNet-50 against FGSM adversarial attacks and payload injection, revealing significant risks in deploying neural networks in security-sensitive environments.

Contribution

It provides a comprehensive analysis of ResNet-50's robustness against specific attack methods and demonstrates the effectiveness of payload injection in manipulating model predictions.

Findings

FGSM attacks increase model confidence in wrong predictions

Payload injection achieves 93.33% success rate in manipulation

Model accuracy remains at 53.33% on clean images

Abstract

This paper investigates the resilience of a ResNet-50 image classification model under two prominent security threats: Fast Gradient Sign Method (FGSM) adversarial attacks and malicious payload injection. Initially, the model attains a 53.33% accuracy on clean images. When subjected to FGSM perturbations, its overall accuracy remains unchanged; however, the model's confidence in incorrect predictions notably increases. Concurrently, a payload injection scheme is successfully executed in 93.33% of the tested samples, revealing how stealthy attacks can manipulate model predictions without degrading visual quality. These findings underscore the vulnerability of even high-performing neural networks and highlight the urgency of developing more robust defense mechanisms for security-critical applications.