| 発表年月日 | 2016/07 |
|---|---|
| 発表テーマ | Suggestive Painterly Style Image Generation System to Satisfy User Preferences |
| 会議名 | ACM SIGGRAPH 2016 |
| 開催地名 | Anaheim |
| 学会区分 | 国際学会 |
| 発表形式 | ポスター |
| 単独共同区分 | 共同 |
| 発表者・共同発表者 | Shuhei Kodama, Tokiichiro Takahashi |
| 概要 | In non-photorealistic rendering (NPR), an evaluated value of the rendered image is not absolute. It is different for each user. Therefore, in NPR, it is important to render an image that satisfies user preferences. Many methods of painterly style image generation have been proposed/developed. However, these methods focus on image generation that imitates painting material or a painterly style. Therefore, to the best of our knowledge, the study of how to generate a painterly style image that satisfies user preferences dose not yet exist. When gazing at an image, the user is more effective at quickly identifying regions that the user dislikes than in finding preferred regions. To address this point, we propose a mechanism of painterly style rendering that satisfies user preferences. |