Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing
The generative adversarial neural network has shown a novel result in the image generation area. However, applying it to a semantic segmentation inpainting task exhibits instability due to the different data distribution. To solve this problem, we propose an unsupervised semantic segmentation inpain...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/2/781 |
| _version_ | 1797446786073231360 |
|---|---|
| author | Woo-Jin Ahn Dong-Won Kim Tae-Koo Kang Dong-Sung Pae Myo-Taeg Lim |
| author_facet | Woo-Jin Ahn Dong-Won Kim Tae-Koo Kang Dong-Sung Pae Myo-Taeg Lim |
| author_sort | Woo-Jin Ahn |
| collection | DOAJ |
| description | The generative adversarial neural network has shown a novel result in the image generation area. However, applying it to a semantic segmentation inpainting task exhibits instability due to the different data distribution. To solve this problem, we propose an unsupervised semantic segmentation inpainting method using an adversarial deep neural network with a newly introduced preprocessing method and loss function. For stabilizing the adversarial training for semantic segmentation inpainting, we match the probability distribution of the segmentation maps with the developed preprocessing method. In addition, a new cross-entropy total variation loss for the probability map is introduced to improve the segmentation inpainting work by smoothing the segmentation map. The experimental results demonstrate the proposed algorithm’s effectiveness on both synthetic and real datasets. |
| first_indexed | 2024-03-09T13:45:35Z |
| format | Article |
| id | doaj.art-c9e563df8e424598a1a8b65a49249126 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T13:45:35Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-c9e563df8e424598a1a8b65a492491262023-11-30T21:01:20ZengMDPI AGApplied Sciences2076-34172023-01-0113278110.3390/app13020781Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with PreprocessingWoo-Jin Ahn0Dong-Won Kim1Tae-Koo Kang2Dong-Sung Pae3Myo-Taeg Lim4School of Electrical Engineering, Korea University, Seoul 02841, Republic of KoreaDepartment of Digital Electronics, Inha Technical College, Incheon 22212, Republic of KoreaDepartment of Human Intelligence and Robot Engineering, Sangmyung University, Cheonan 31066, Republic of KoreaDepartment of Software, Sangmyung University, Cheonan 31066, Republic of KoreaSchool of Electrical Engineering, Korea University, Seoul 02841, Republic of KoreaThe generative adversarial neural network has shown a novel result in the image generation area. However, applying it to a semantic segmentation inpainting task exhibits instability due to the different data distribution. To solve this problem, we propose an unsupervised semantic segmentation inpainting method using an adversarial deep neural network with a newly introduced preprocessing method and loss function. For stabilizing the adversarial training for semantic segmentation inpainting, we match the probability distribution of the segmentation maps with the developed preprocessing method. In addition, a new cross-entropy total variation loss for the probability map is introduced to improve the segmentation inpainting work by smoothing the segmentation map. The experimental results demonstrate the proposed algorithm’s effectiveness on both synthetic and real datasets.https://www.mdpi.com/2076-3417/13/2/781deeplearningconvolutional neural networkdata preprocessingbinary total variation |
| spellingShingle | Woo-Jin Ahn Dong-Won Kim Tae-Koo Kang Dong-Sung Pae Myo-Taeg Lim Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing Applied Sciences deeplearning convolutional neural network data preprocessing binary total variation |
| title | Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing |
| title_full | Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing |
| title_fullStr | Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing |
| title_full_unstemmed | Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing |
| title_short | Unsupervised Semantic Segmentation Inpainting Network Using a Generative Adversarial Network with Preprocessing |
| title_sort | unsupervised semantic segmentation inpainting network using a generative adversarial network with preprocessing |
| topic | deeplearning convolutional neural network data preprocessing binary total variation |
| url | https://www.mdpi.com/2076-3417/13/2/781 |
| work_keys_str_mv | AT woojinahn unsupervisedsemanticsegmentationinpaintingnetworkusingagenerativeadversarialnetworkwithpreprocessing AT dongwonkim unsupervisedsemanticsegmentationinpaintingnetworkusingagenerativeadversarialnetworkwithpreprocessing AT taekookang unsupervisedsemanticsegmentationinpaintingnetworkusingagenerativeadversarialnetworkwithpreprocessing AT dongsungpae unsupervisedsemanticsegmentationinpaintingnetworkusingagenerativeadversarialnetworkwithpreprocessing AT myotaeglim unsupervisedsemanticsegmentationinpaintingnetworkusingagenerativeadversarialnetworkwithpreprocessing |