Attribution for Enhanced Explanation with Transferable Adversarial eXploration
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929648872456192 |
|---|---|
| author | Zhu, Zhiyu Zhang, Jiayu Jin, Zhibo Chen, Huaming Zhou, Jianlong Chen, Fang |
| author_facet | Zhu, Zhiyu Zhang, Jiayu Jin, Zhibo Chen, Huaming Zhou, Jianlong Chen, Fang |
| contents | 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/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_19523 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Attribution for Enhanced Explanation with Transferable Adversarial eXploration Zhu, Zhiyu Zhang, Jiayu Jin, Zhibo Chen, Huaming Zhou, Jianlong Chen, Fang Artificial Intelligence Computer Vision and Pattern Recognition 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/ |
| title | Attribution for Enhanced Explanation with Transferable Adversarial eXploration |
| topic | Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.19523 |