Regional Patch-Based Feature Interpolation Method for Effective Regularization
Deep Convolutional Neural Networks (CNNs) can be overly dependent on training data, causing a generalization problem in which trained models may not predict real-world datasets. To address this problem, various regularization methods such as image manipulation and feature map regularization have bee...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9000503/ |
| _version_ | 1818854807247519744 |
|---|---|
| author | Soojin Jang Kyohoon Jin Junhyeok An Youngbin Kim |
| author_facet | Soojin Jang Kyohoon Jin Junhyeok An Youngbin Kim |
| author_sort | Soojin Jang |
| collection | DOAJ |
| description | Deep Convolutional Neural Networks (CNNs) can be overly dependent on training data, causing a generalization problem in which trained models may not predict real-world datasets. To address this problem, various regularization methods such as image manipulation and feature map regularization have been proposed for their strong generalization ability. In this paper, we propose a regularization method that applies both image manipulation and feature map regularization based on patches. The method proposed in this paper has a regularization effect in two stages, which makes it possible to better generalize the model. Consequently, it improves the performance of the model. Moreover, our method adds features extracted from other images in the hidden state stage, which not only makes the model robust to noise but also captures the distribution of each label. Through experiments, we show that our method performs competently on models that generate a large number of parameter and multiple feature maps for the CIFAR and Tiny-ImageNet datasets. |
| first_indexed | 2024-12-19T07:58:34Z |
| format | Article |
| id | doaj.art-18c82393b6534d6da1cf61ccc26f6bde |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-19T07:58:34Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-18c82393b6534d6da1cf61ccc26f6bde2022-12-21T20:29:55ZengIEEEIEEE Access2169-35362020-01-018336583366510.1109/ACCESS.2020.29743289000503Regional Patch-Based Feature Interpolation Method for Effective RegularizationSoojin Jang0Kyohoon Jin1Junhyeok An2Youngbin Kim3https://orcid.org/0000-0002-2114-0120Department of Image Science and Arts, Chung-Ang University, Seoul, South KoreaDepartment of Image Science and Arts, Chung-Ang University, Seoul, South KoreaDepartment of Image Science and Arts, Chung-Ang University, Seoul, South KoreaDepartment of Image Science and Arts, Chung-Ang University, Seoul, South KoreaDeep Convolutional Neural Networks (CNNs) can be overly dependent on training data, causing a generalization problem in which trained models may not predict real-world datasets. To address this problem, various regularization methods such as image manipulation and feature map regularization have been proposed for their strong generalization ability. In this paper, we propose a regularization method that applies both image manipulation and feature map regularization based on patches. The method proposed in this paper has a regularization effect in two stages, which makes it possible to better generalize the model. Consequently, it improves the performance of the model. Moreover, our method adds features extracted from other images in the hidden state stage, which not only makes the model robust to noise but also captures the distribution of each label. Through experiments, we show that our method performs competently on models that generate a large number of parameter and multiple feature maps for the CIFAR and Tiny-ImageNet datasets.https://ieeexplore.ieee.org/document/9000503/Convolutional neural networkmanifoldregularizationcomputer vision |
| spellingShingle | Soojin Jang Kyohoon Jin Junhyeok An Youngbin Kim Regional Patch-Based Feature Interpolation Method for Effective Regularization IEEE Access Convolutional neural network manifold regularization computer vision |
| title | Regional Patch-Based Feature Interpolation Method for Effective Regularization |
| title_full | Regional Patch-Based Feature Interpolation Method for Effective Regularization |
| title_fullStr | Regional Patch-Based Feature Interpolation Method for Effective Regularization |
| title_full_unstemmed | Regional Patch-Based Feature Interpolation Method for Effective Regularization |
| title_short | Regional Patch-Based Feature Interpolation Method for Effective Regularization |
| title_sort | regional patch based feature interpolation method for effective regularization |
| topic | Convolutional neural network manifold regularization computer vision |
| url | https://ieeexplore.ieee.org/document/9000503/ |
| work_keys_str_mv | AT soojinjang regionalpatchbasedfeatureinterpolationmethodforeffectiveregularization AT kyohoonjin regionalpatchbasedfeatureinterpolationmethodforeffectiveregularization AT junhyeokan regionalpatchbasedfeatureinterpolationmethodforeffectiveregularization AT youngbinkim regionalpatchbasedfeatureinterpolationmethodforeffectiveregularization |