Acknowledgement
This work was supported by Electronics and Telecommunications Research Institute (ETRI) grant funded by the Korean government. [22ZH1200, The research of the fundamental media·contents technologies for hyper-realistic media space].
Annual Conference of KIPS (한국정보처리학회:학술대회논문집)
- 2022.11a
- /
- Pages.648-649
- /
- 2022
- /
- 2005-0011(pISSN)
- /
- 2671-7298(eISSN)
DOI QR Code
Transformer-based glass area detection method
트랜스포머기 반 유리 영역 검출방법
- Hu, Xiaohang (Dept of Multimedia Engineering, Dongguk University) ;
- Gao, Rui (Dept of Multimedia Engineering, Dongguk University) ;
- Yang, Seung-Jun (Electronics and Telecommunications Research Institute) ;
- Cho, Kyungeun (Dept of Multimedia Engineering, Dongguk University)
- Published : 2022.11.21
Abstract
Glass is a common object in living environments, but even humans are sometimes unable to identify it. This study proposes a method for detecting glass area by learning edge information from images. The network structure of Transformer is used to accept the base features extracted by backbone and extract the boundary information of RGB images, and both features are used to learn the features of glass area and determine the glass area based on these boundary features. The experimental results show that our proposed method can detect glass area in images.
Keywords