V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation
Infrared image simulation is challenging because it is complex to model. To estimate the corresponding infrared image directly from the visible light image, we propose a three-level refined light-weight generative adversarial network with cascaded guidance (V2T-GAN), which can improve the accuracy o...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/6/2119 |
| _version_ | 1827625730270298112 |
|---|---|
| author | Ruiming Jia Xin Chen Tong Li Jiali Cui |
| author_facet | Ruiming Jia Xin Chen Tong Li Jiali Cui |
| author_sort | Ruiming Jia |
| collection | DOAJ |
| description | Infrared image simulation is challenging because it is complex to model. To estimate the corresponding infrared image directly from the visible light image, we propose a three-level refined light-weight generative adversarial network with cascaded guidance (V2T-GAN), which can improve the accuracy of the infrared simulation image. V2T-GAN is guided by cascading auxiliary tasks and auxiliary information: the first-level adversarial network uses semantic segmentation as an auxiliary task, focusing on the structural information of the infrared image; the second-level adversarial network uses the grayscale inverted visible image as the auxiliary task to supplement the texture details of the infrared image; the third-level network obtains a sharp and accurate edge by adding auxiliary information of the edge image and a displacement network. Experiments on the public dataset Multispectral Pedestrian Dataset demonstrate that the structure and texture features of the infrared simulation image obtained by V2T-GAN are correct, and outperform the state-of-the-art methods in objective metrics and subjective visualization effects. |
| first_indexed | 2024-03-09T12:42:42Z |
| format | Article |
| id | doaj.art-49acd481618d46b69c541f873691ed3c |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T12:42:42Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-49acd481618d46b69c541f873691ed3c2023-11-30T22:16:30ZengMDPI AGSensors1424-82202022-03-01226211910.3390/s22062119V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal TranslationRuiming Jia0Xin Chen1Tong Li2Jiali Cui3School of Information Science and Technology, North China University of Technology, Beijing 100144, ChinaSchool of Information Science and Technology, North China University of Technology, Beijing 100144, ChinaSchool of Information Science and Technology, North China University of Technology, Beijing 100144, ChinaSchool of Information Science and Technology, North China University of Technology, Beijing 100144, ChinaInfrared image simulation is challenging because it is complex to model. To estimate the corresponding infrared image directly from the visible light image, we propose a three-level refined light-weight generative adversarial network with cascaded guidance (V2T-GAN), which can improve the accuracy of the infrared simulation image. V2T-GAN is guided by cascading auxiliary tasks and auxiliary information: the first-level adversarial network uses semantic segmentation as an auxiliary task, focusing on the structural information of the infrared image; the second-level adversarial network uses the grayscale inverted visible image as the auxiliary task to supplement the texture details of the infrared image; the third-level network obtains a sharp and accurate edge by adding auxiliary information of the edge image and a displacement network. Experiments on the public dataset Multispectral Pedestrian Dataset demonstrate that the structure and texture features of the infrared simulation image obtained by V2T-GAN are correct, and outperform the state-of-the-art methods in objective metrics and subjective visualization effects.https://www.mdpi.com/1424-8220/22/6/2119image domain translationinfrared image simulationgenerative adversarial network |
| spellingShingle | Ruiming Jia Xin Chen Tong Li Jiali Cui V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation Sensors image domain translation infrared image simulation generative adversarial network |
| title | V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation |
| title_full | V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation |
| title_fullStr | V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation |
| title_full_unstemmed | V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation |
| title_short | V2T-GAN: Three-Level Refined Light-Weight GAN with Cascaded Guidance for Visible-to-Thermal Translation |
| title_sort | v2t gan three level refined light weight gan with cascaded guidance for visible to thermal translation |
| topic | image domain translation infrared image simulation generative adversarial network |
| url | https://www.mdpi.com/1424-8220/22/6/2119 |
| work_keys_str_mv | AT ruimingjia v2tganthreelevelrefinedlightweightganwithcascadedguidanceforvisibletothermaltranslation AT xinchen v2tganthreelevelrefinedlightweightganwithcascadedguidanceforvisibletothermaltranslation AT tongli v2tganthreelevelrefinedlightweightganwithcascadedguidanceforvisibletothermaltranslation AT jialicui v2tganthreelevelrefinedlightweightganwithcascadedguidanceforvisibletothermaltranslation |