-
Article type: Cover
Pages
Cover1-
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Index
Pages
Toc1-
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Yuki Hashimoto, Daigo Muramatsu, Hiroyuki Ogata
Article type: Article
Session ID: ME2010-164
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
The manner of holding a pen is distinctive among people. Therefore, pen holding style is useful for person authentication. In this paper, we propose a biometric person authentication method using features extracted from images of pen holding style. Images of the pen holding style are captured by a camera, and several features are extracted from the captured images. These features are compared with a reference dataset to calculate dissimilarity scores, and these scores are combined for verification using a three-layer perceptron. Preliminary experiments were performed by using a private database. The proposed system yielded an equal error rate (EER) of 2.6% against an impersonate attack..
View full abstract
-
Seiichiro HANGAI
Article type: Article
Session ID: ME2010-165
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In ISO/IEC JTC1 SC37 WG3, a couple of signature/sign data formats are discussed and on the way to standardization. In this article, I will overview both formats and point out problems in applying them to Japanese signature/sign recording and verification.
View full abstract
-
Kyohei KOYAMA, Takahiro YOSHIDA, Naoya WADA, Seiichiro HANGAI
Article type: Article
Session ID: ME2010-166
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In online signature verification, an improvement of robustness to imitating forgeries and tracing forgeries is subject to consider for installation. In other word, it is important to reject such forgeries efficiently. In this report, we have investigated a screening method using mean signature time and mean pen altitude from the view point of the rejection rate of two kinds of forgeries written by 20 persons. From the result, we can reject 85% forgeries only by mean signature time and more than 50% forgeries by mean pen altitude.
View full abstract
-
[in Japanese]
Article type: Article
Session ID: ME2010-167
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Tomohiro Yamamoto, Kenta Mizui, Makoto Okabe, Rikio Onai
Article type: Article
Session ID: ME2010-168
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We propose the method to automatically synthesize a video, where a player looks like playing a user-specified music. Our current target performance is violin solo. We first create a video database of violin solo performance. For each video, we analyze its music and extract the timing of each note. Given an input music, we divide it into fragments and analyze timings of notes. For each fragment, the timing information is used to search for an appropriate video from the database. We render the final animation applying time warping. We demonstrate our method can synthesize a fake performance video of various types of violin solo music.
View full abstract
-
Hiroya KITADA, Kazuhito MURAKAMI
Article type: Article
Session ID: ME2010-169
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper proposes a thumbnail generation method based on scale-space analysis. First, scale-space analysis is applied to digital flows of the image in order to obtain reasonable scale. Then, some candidates of thumbnail are extracted based on the result of sub-regions calculated by the digital flows. Finally, the similarity between each candidate of thumbnail is calculated by using the distribution of brightness. Some experimental results show the effectiveness of our proposed method.
View full abstract
-
Mark Duppenthaler, Ken Tsutsuguchi, Akira Kojima
Article type: Article
Session ID: ME2010-170
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We describe our image vectorization and editing system. Using a technique based on Xia et al. [8], our vectorization takes a bitmap image and uses a triangular mesh to topologically represent it. The vectorization is fully automatic; in addition, the user can manually select relevant objects which can be individually vectorized. The color inside each patch is represented using thin plate spline surfaces. Each vectorized image is stored as XML and can be loaded on web browsers for displaying and editing. Patches can be tagged with names, moved or stretched, and the interior color can be changed at a more abstract level compared to editing individual pixels.
View full abstract
-
Hiroki NINOMIYA, Kazuhito MURAKAMI
Article type: Article
Session ID: ME2010-171
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
The distribution of facial temperature depends on the person and the region that has higher temperature is stable against the change of environment temperature. This paper proposes a method to recognize individual face based on the distribution of facial temperature. This method utilizes the normalized cross-correlation for individual recognition. Experimental results show that the recognition rate is 85.4%.
View full abstract
-
Shoko YONEZAWA, Manabu INUMA, Akira OTSUKA, Hideki IMAI
Article type: Article
Session ID: ME2010-172
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We classify the attacks against biometric authentication systems by attacker's capability and define the security against impersonation attacks. We also map the known attacks such as test object attacks and wolf attacks to this classification, and clarify the abilities and goals of these attacks. Moreover, we point out the potential attacks against biometric authentication systems according to our classification and discuss the desired security of biometric authentication systems.
View full abstract
-
Takumi YAMAMOTO, Tokuichiro SUZUKI, J. D. Tygar, Masakatsu NISHIGAKI
Article type: Article
Session ID: ME2010-173
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
Recently as many researchers have reported, the conventional CAPTCHA could be defeated by recent malwares since the ability of PCs get closer to that of human. Therefore CAPTCHA should be based on an even more advanced human cognitive processing ability. To realize a new CAPTCHA, we have already focused on human abilities "to recognize strangeness" and "to understand humor", and proposed 2-types of new CAPTCHA which utilize those abilities. In this paper, we briefly introduce these 2 CAPTCHAs and have discussion on usability, security and operability in practical use.
View full abstract
-
Daishi Watabe, Koji Soma, Zichong Huang, Hideyasu Sai, Katsuhiro Sakai ...
Article type: Article
Session ID: ME2010-174
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
A surveillance system based on ear biometrics is discussed.
View full abstract
-
Junpei SAGARA, Kazuhito MURAKAMI
Article type: Article
Session ID: ME2010-175
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper clarifies the relation between the saliency of hue, saturation and brightness of original image and the "conspicuousness" of the selected region, and discusses what kind of color feature makes a contribution to the conspicuousness of the region. First, the saliencies of hue, saturation and brightness are calculated in the image by subtracting different scale images. Then, the most salient region for each feature is extracted by moving the region of interest. In the experiment, we examine the effectiveness of our method. By the result of experiment, it was clarified that estimating conspicuous region in the image was possible from saliencies of each element and image features.
View full abstract
-
Yuri JITSUKATA, Kan OKUBO, Norio TAGAWA
Article type: Article
Session ID: ME2010-176
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
Small vibrations of eyeball, which occur when we fix our gaze on object, are called "fixational eye movements." We have studied 3-D depth recovery methods using the camera model imitating such the eye movements. These methods are defined as a differential form, i.e. spatio-temporal derivatives of successive two images are analyzed. However, in these methods, observed noise of derivatives causes serious recovering error. In this study, we examine a new method using the same camera model, in which local image blurring including depth information is analyzed.
View full abstract
-
Todorka Alexandrova, Yoshiaki Suzuki, Kan Okubo, Norio Tagawa
Article type: Article
Session ID: ME2010-177
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper a method for realizing a secret images sharing scheme with general access structure has been proposed. The proposed algorithm for secret images sharing is realized by applying multi-secret sharing schemes based on two-variable one-way functions.
View full abstract
-
Article type: Appendix
Pages
App1-
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Appendix
Pages
App2-
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Appendix
Pages
App3-
Published: December 09, 2010
Released on J-STAGE: September 21, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS