Research on Automatic Digest Generation for Comic Retrieval
| Project/Area Number |
25330137
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Multimedia database
|
|---|
| Research Institution | Waseda University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
ISHI Daisuke 早稲田大学, 理工学術院, 助手 (40581525)
|
|---|
| Project Period (FY) |
2013-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥4,420,000 (Direct Cost: ¥3,400,000、Indirect Cost: ¥1,020,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2014: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2013: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
|
|---|
| Keywords | マンガ / 顔画像検出 / 機械学習 / 自動要約 / キャラクター抽出 / HOG特徴量 / Deformable Part Model / ロールコミュニティー / マンガ画像 / 顔画像認識 / Haar-like特徴量 / キャラクター認識 |
|---|
| Outline of Final Research Achievements |
It is important to extract metadata contained in comics for realizing automatic digest generation system. The most important metadata is comic characters. In this research, we focused on the improvement of comic character extraction. Through experiments, we showed that Deformable Part Model (DPM), which is based on HOG features, is effective for extraction. We also studied on main character identification without prior knowledge by clustering. By this approach, positive and negative samples can be fed into DPM. Further, "Role Community Model", which is the method to identify the weight of each page by counting locations of character's appearance and their relations, can be used to realize automatic comic digest.
|
Report
(4 results)
Research Products
(16 results)
