Abstract | ||
---|---|---|
In this paper, we study realistic bokeh rendering from a single all-in-focus image. Existing computational bokeh rendering methods generate bokeh effects by adding a simple flat background blur. As a result, the rendering results are different from the real bokeh on DSLR cameras. To address this issue, we propose a multi-stage network to learn shallow depth-of-field from a single bokeh-free image. In particular, our network consists of four modules: defocus estimation, radiance, rendering, and upsampling. The four modules are trained on different sizes to learn global features as well as local details around the boundaries of in-focus objects. Experimental results show that our approach is capable of rendering a pleasing distinctive bokeh effect in complex scenes. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1007/978-3-030-67070-2_15 | ECCV Workshops |
Keywords | DocType | Citations |
Bokeh rendering,Defocus estimation,Radiance,Upsampling | Conference | 2 |
PageRank | References | Authors |
0.38 | 2 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xianrui Luo | 1 | 5 | 1.77 |
Juewen Peng | 2 | 5 | 1.10 |
Ke Xian | 3 | 55 | 8.99 |
Zijin Wu | 4 | 5 | 0.76 |
Zhiguo Cao | 5 | 314 | 44.17 |