1. NCCU Academic Hub
  2. 學術產出
  3. 資訊學院
  4. 會議論文
  5. Item

學術產出-會議論文

文章檢視/開啟

書目匯出

Google ScholarTM

政大圖書館

引文資訊

TAIR相關學術產出

題名 Rain2Avoid: Self-Supervised Single Image Deraining
作者 彭彥璁
Peng, Yan-Tsung;Li, Wei-Hua
貢獻者 資訊系
關鍵詞 Image deraining; self-supervised; stochastic derained references
日期 2023-06
上傳時間 16-二月-2024 15:36:56 (UTC+8)
摘要 The single image deraining task aims to remove rain from a single image, attracting much attention in the field. Recent research on this topic primarily focuses on discriminative deep learning methods, which train models on rainy images with their clean counterparts. However, collecting such paired images for training takes much work. Thus, we present Rain2Avoid (R2A), a training scheme that requires only rainy images for image deraining. We propose a locally dominant gradient prior to reveal possible rain streaks and overlook those rain pixels while training with the input rainy image directly. Understandably, R2A may not perform as well as deraining methods that supervise their models with rain-free ground truth. However, R2A favors when training image pairs are unavailable and can self-supervise only one rainy image for deraining. Experimental results show that the proposed method performs favorably against state-of-the-art few-shot deraining and self-supervised denoising methods.
關聯 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE
資料類型 conference
DOI https://doi.org/10.1109/ICASSP49357.2023.10097092
dc.contributor 資訊系
dc.creator (作者) 彭彥璁
dc.creator (作者) Peng, Yan-Tsung;Li, Wei-Hua
dc.date (日期) 2023-06
dc.date.accessioned 16-二月-2024 15:36:56 (UTC+8)-
dc.date.available 16-二月-2024 15:36:56 (UTC+8)-
dc.date.issued (上傳時間) 16-二月-2024 15:36:56 (UTC+8)-
dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/149884-
dc.description.abstract (摘要) The single image deraining task aims to remove rain from a single image, attracting much attention in the field. Recent research on this topic primarily focuses on discriminative deep learning methods, which train models on rainy images with their clean counterparts. However, collecting such paired images for training takes much work. Thus, we present Rain2Avoid (R2A), a training scheme that requires only rainy images for image deraining. We propose a locally dominant gradient prior to reveal possible rain streaks and overlook those rain pixels while training with the input rainy image directly. Understandably, R2A may not perform as well as deraining methods that supervise their models with rain-free ground truth. However, R2A favors when training image pairs are unavailable and can self-supervise only one rainy image for deraining. Experimental results show that the proposed method performs favorably against state-of-the-art few-shot deraining and self-supervised denoising methods.
dc.format.extent 113 bytes-
dc.format.mimetype text/html-
dc.relation (關聯) IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE
dc.subject (關鍵詞) Image deraining; self-supervised; stochastic derained references
dc.title (題名) Rain2Avoid: Self-Supervised Single Image Deraining
dc.type (資料類型) conference
dc.identifier.doi (DOI) 10.1109/ICASSP49357.2023.10097092
dc.doi.uri (DOI) https://doi.org/10.1109/ICASSP49357.2023.10097092