| Presentation | 2016-03-10 Feature Point Extraction from A Manga-Style Illustration of A Facial Image Using SIFT and Color Features as A Feature Vector Shun Aoki, Ryusuke Miyamoto, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Several schemes have been proposed to localize facial parts in natural imamges accurately because many practical applications requires accurately extracted facial keypoints. It is very difficult but a novel approach hasshown a good accuracy for this task: SIFT for local feature extraction and RANSAC-like algorithm for estimation of a global structure. However, this scheme does not work well on illustrations and images including large regions with nearly solid color that is often obtained from animation and 3DGC. To improve the estimation accuracy of facial parts in facial illustrations, we propose a novel scheme that uses color features in addition to SIFT features in local feature extraction of the existing scheme. In the proposed scheme, a feature vector of a local feature extractor consists of a SIFT feature from a image resolution and color histograms though the original scheme uses two SIFT features from two image resolutions. To evaluate the proposed scheme, we have compared the proposed scheme and the existing scheme using a dataset that consits of 1500 images with the coordinates of 13 feature points for a face created by the authors. Experimental results show that the proposed scheme can reduce the estimation error to 41.80% of the inter-ocular distance if the color space, the window size, and the shape of local feature extractor areLUV, 11×11, and a square, respectively though the existing scheme shows 76.77% of the inter-ocular distance. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | feature point localization / facial image / animation character / SIFT / color histogram |
| Paper # | SIS2015-59 |
| Date of Issue | 2016-03-03 (SIS) |
| Conference Information | |
| Committee | SIS |
|---|---|
| Conference Date | 2016/3/10(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Tokyo City Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Soft computing, etc. |
| Chair | Mitsuji Muneyasu(Kansai Univ.) |
| Vice Chair | Hirokazu Tanaka(Hiroshima City Univ.) / Takayuki Nakachi(NTT) |
| Secretary | Hirokazu Tanaka(Nagoya City Univ.) / Takayuki Nakachi(Toshiba) |
| Assistant | Hiroyuki Tsuji(Kanagawa Inst. of Tech.) / Hakaru Tamukoh(Kyushu Inst. of Tech.) |
| Paper Information | |
| Registration To | Technical Committee on Smart Info-Media System |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Feature Point Extraction from A Manga-Style Illustration of A Facial Image Using SIFT and Color Features as A Feature Vector |
| Sub Title (in English) | |
| Keyword(1) | feature point localization |
| Keyword(2) | facial image |
| Keyword(3) | animation character |
| Keyword(4) | SIFT |
| Keyword(5) | color histogram |
| 1st Author's Name | Shun Aoki |
| 1st Author's Affiliation | Meiji University(Meiji Univ.) |
| 2nd Author's Name | Ryusuke Miyamoto |
| 2nd Author's Affiliation | Meiji University(Meiji Univ.) |
| Date | 2016-03-10 |
| Paper # | SIS2015-59 |
| Volume (vol) | vol.115 |
| Number (no) | SIS-505 |
| Page | pp.pp.63-68(SIS), |
| #Pages | 6 |
| Date of Issue | 2016-03-03 (SIS) |