Adversarial Attacks Assessment of Salient Object Detection via Symbolic Learning

TL;DR

This paper demonstrates that symbolic learning via brain programming offers superior robustness against adversarial attacks in salient object detection compared to deep neural networks, which are vulnerable to perturbations.

Contribution

It provides empirical evidence that symbolic learning methods are more resilient to adversarial attacks in visual attention tasks than traditional deep learning models.

Findings

Neural networks suffer significant performance drops under attacks.

Brain programming remains unaffected by intense perturbations.

Symbolic learning enhances reliability in security-critical applications.

Abstract

Machine learning is at the center of mainstream technology and outperforms classical approaches to handcrafted feature design. Aside from its learning process for artificial feature extraction, it has an end-to-end paradigm from input to output, reaching outstandingly accurate results. However, security concerns about its robustness to malicious and imperceptible perturbations have drawn attention since its prediction can be changed entirely. Salient object detection is a research area where deep convolutional neural networks have proven effective but whose trustworthiness represents a significant issue requiring analysis and solutions to hackers' attacks. Brain programming is a kind of symbolic learning in the vein of good old-fashioned artificial intelligence. This work provides evidence that symbolic learning robustness is crucial in designing reliable visual attention systems since it can withstand even the most intense perturbations. We test this evolutionary computation methodology against several adversarial attacks and noise perturbations using standard databases and a real-world problem of a shorebird called the Snowy Plover portraying a visual attention task. We compare our methodology with five different deep learning approaches, proving that they do not match the symbolic paradigm regarding robustness. All neural networks suffer significant performance losses, while brain programming stands its ground and remains unaffected. Also, by studying the Snowy Plover, we remark on the importance of security in surveillance activities regarding wildlife protection and conservation.