Edge-Informed Single Image Super-Resolution
這篇文章提出了一種將圖像結構信息和紋理信息分開處理的單圖像超分辨率方法，通過這種方法將單圖像超分辨率轉換爲圖像修復。 該文章首先分析了傳統的雙三次插值的弊端： 原標準的HR圖像是有邊緣信息的高清圖像，2倍下采樣時四個像素合成一個像素點使下采樣後的圖片丟失了高頻邊緣的結構信息和方向信息，後雙三次插值得到的2倍恢復的原圖徹底丟失了高頻邊緣的結構信息和方向信息。 該文章提出了一種，即通過插值法將圖像放大
相關文章
- 1. 【超分辨率】Enhanced Deep Residual Networks for Single Image Super-Resolution
- 2. DeeperLab: Single-Shot Image Parser 詳解
- 3. 3D Face Reconstruction from a Single Image
- 4. AIPNet: Image-to-Image Single Image Dehazing with Atmospheric Illumination Prior
- 5. GANimation: Anatomically-aware Facial Animation from a Single Image
- 6. RAM: Residual Attention Module for Single Image Super-Resolution
- 7. SinGAN：Learning a Generative Model from a Single Natural Image
- 8. Depth-attentional Features for Single-image Rain Removal
- 9. GANimation: Anatomically-aware Facial Animation from a Single Image（ECCV18）
- 10. A Cascaded Convolutional Neural Network for Single Image Dehazing
- 更多相關文章...
- • RSS
元素 - RSS 教程 - • ASP.NET Image 控件 - ASP.NET 教程
- • RxJava操作符（十）自定義操作符
- • Docker 清理命令
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. 【超分辨率】Enhanced Deep Residual Networks for Single Image Super-Resolution
- 2. DeeperLab: Single-Shot Image Parser 詳解
- 3. 3D Face Reconstruction from a Single Image
- 4. AIPNet: Image-to-Image Single Image Dehazing with Atmospheric Illumination Prior
- 5. GANimation: Anatomically-aware Facial Animation from a Single Image
- 6. RAM: Residual Attention Module for Single Image Super-Resolution
- 7. SinGAN：Learning a Generative Model from a Single Natural Image
- 8. Depth-attentional Features for Single-image Rain Removal
- 9. GANimation: Anatomically-aware Facial Animation from a Single Image（ECCV18）
- 10. A Cascaded Convolutional Neural Network for Single Image Dehazing