Semantic Information Supplementary Pyramid Network for Dynamic Scene Deblurring
| Autor: | Duqiang Luo, Ying Qiao, Dahong Xu, Junhui Li, Yiming Liu, Wenzhuo Huang, Yifei Luo |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Deblurring
feature pyramid network General Computer Science Computer science Feature extraction 02 engineering and technology Semantics semantic information Pyramid 0202 electrical engineering electronic engineering information engineering Code (cryptography) General Materials Science Pyramid (image processing) Image restoration business.industry General Engineering 020206 networking & telecommunications Pattern recognition Image segmentation Feature (computer vision) 020201 artificial intelligence & image processing lcsh:Electrical engineering. Electronics. Nuclear engineering Artificial intelligence Generative adversarial network business lcsh:TK1-9971 |
| Zdroj: | IEEE Access, Vol 8, Pp 188587-188599 (2020) |
| ISSN: | 2169-3536 |
| Popis: | The algorithm in this paper is called semantic information supplementary pyramid network(SIS-net). We choose Generative Adversarial Network (GAN) as its fundamental model. SIS-net's generator imitates the feature pyramid network (FPN) structure to recycle features spanning across multiple receptive scales to restore a sharp image. However, to solve the problem caused by the phenomenon of semantic dilution in the FPN network, we have innovatively designed a semantic information supplement (SIS) mechanism. SIS mechanism contains two essential components: semantic information storage box (info-box) and feature fusion expanding. In the process of feature fusion expanding, the semantic information features coming from the info-box is supplemented to make greater use of detailed clues. In addition, SIS-net uses the intermediate layer path to extract image features in a single time to obtain a multi-scale effect. The running speed of SIS-net has obvious advantages over other algorithms, and can basically complete real-time deblurring tasks. Extensive experiments show that our SIS-net achieves both qualitative and quantitative improvements against state-of-the-art methods. The code is available at https://github.com/yimingliu123/SIS-NET. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|