Understanding CNNs from excitations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2022
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913194571726848 |
|---|---|
| author | Ying, Zijian Li, Qianmu Lian, Zhichao Hou, Jun Lin, Tong Wang, Tao |
| author_facet | Ying, Zijian Li, Qianmu Lian, Zhichao Hou, Jun Lin, Tong Wang, Tao |
| contents | Saliency maps have proven to be a highly efficacious approach for explicating the decisions of Convolutional Neural Networks. However, extant methodologies predominantly rely on gradients, which constrain their ability to explicate complex models. Furthermore, such approaches are not fully adept at leveraging negative gradient information to improve interpretive veracity. In this study, we present a novel concept, termed positive and negative excitation, which enables the direct extraction of positive and negative excitation for each layer, thus enabling complete layer-by-layer information utilization sans gradients. To organize these excitations into final saliency maps, we introduce a double-chain backpropagation procedure. A comprehensive experimental evaluation, encompassing both binary classification and multi-classification tasks, was conducted to gauge the effectiveness of our proposed method. Encouragingly, the results evince that our approach offers a significant improvement over the state-of-the-art methods in terms of salient pixel removal, minor pixel removal, and inconspicuous adversarial perturbation generation guidance. Additionally, we verify the correlation between positive and negative excitations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2205_00932 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Understanding CNNs from excitations Ying, Zijian Li, Qianmu Lian, Zhichao Hou, Jun Lin, Tong Wang, Tao Computer Vision and Pattern Recognition Machine Learning Saliency maps have proven to be a highly efficacious approach for explicating the decisions of Convolutional Neural Networks. However, extant methodologies predominantly rely on gradients, which constrain their ability to explicate complex models. Furthermore, such approaches are not fully adept at leveraging negative gradient information to improve interpretive veracity. In this study, we present a novel concept, termed positive and negative excitation, which enables the direct extraction of positive and negative excitation for each layer, thus enabling complete layer-by-layer information utilization sans gradients. To organize these excitations into final saliency maps, we introduce a double-chain backpropagation procedure. A comprehensive experimental evaluation, encompassing both binary classification and multi-classification tasks, was conducted to gauge the effectiveness of our proposed method. Encouragingly, the results evince that our approach offers a significant improvement over the state-of-the-art methods in terms of salient pixel removal, minor pixel removal, and inconspicuous adversarial perturbation generation guidance. Additionally, we verify the correlation between positive and negative excitations. |
| title | Understanding CNNs from excitations |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2205.00932 |