Attribution for Enhanced Explanation with Transferable Adversarial eXploration
This work addresses the need for more reliable and stable model explanations in applications like computer vision, though it is incremental as it builds upon an existing framework.
The paper tackles the problem of improving attribution methods for explaining deep neural network decisions in computer vision by introducing AttEXplore++, which enhances accuracy and robustness through transferable adversarial attacks, achieving an average performance improvement of 7.57% over its predecessor and 32.62% compared to other state-of-the-art methods.
The interpretability of deep neural networks is crucial for understanding model decisions in various applications, including computer vision. AttEXplore++, an advanced framework built upon AttEXplore, enhances attribution by incorporating transferable adversarial attack methods such as MIG and GRA, significantly improving the accuracy and robustness of model explanations. We conduct extensive experiments on five models, including CNNs (Inception-v3, ResNet-50, VGG16) and vision transformers (MaxViT-T, ViT-B/16), using the ImageNet dataset. Our method achieves an average performance improvement of 7.57\% over AttEXplore and 32.62\% compared to other state-of-the-art interpretability algorithms. Using insertion and deletion scores as evaluation metrics, we show that adversarial transferability plays a vital role in enhancing attribution results. Furthermore, we explore the impact of randomness, perturbation rate, noise amplitude, and diversity probability on attribution performance, demonstrating that AttEXplore++ provides more stable and reliable explanations across various models. We release our code at: https://anonymous.4open.science/r/ATTEXPLOREP-8435/