|
Affiliation |
Engineering educational research section Information and Communication Technology Program |
USUZAKI Syotaro
|
|
|
Papers 【 display / non-display 】
-
Tsunami evacuation simulation considering road width in Aoshima district Reviewed
Nabeyama K., Usuzaki S., Aburada K., Yamaba H., Katayama T., Okazaki N.
Artificial Life and Robotics 28 ( 4 ) 779 - 788 2023.11
Publishing type：Research paper (scientific journal) Publisher：Artificial Life and Robotics
Japan is one of the countries in the world where natural disasters occur most frequently. Typhoons, earthquakes, tsunamis, volcanic eruptions, and many other natural disasters occur in Japan every year, and the risks are immeasurable. Japan is especially prone to earthquakes because of its location on the North American Plate, Eurasian Plate, Pacific Plate, and Philippine Sea Plate. Preparation for a Nankai Trough earthquake is one of the most important issues. A Nankai trough earthquake is predicted to occur with a probability of 70–80% within 30 years. Accordingly, a tsunami of more than 10 ms is expected to hit a wide area along the Pacific coast from the Kanto region to the Kyushu region, and the impact is expected to be significant. When a tsunami warning is issued, an evacuation support system that provides information on evacuation sites and routes enables faster and safer evacuation actions. In a previous study conducted in our laboratory, we proposed an evacuation support system using Low-Power Wide-Area communications and conducted a simulation to find ways to improve the evacuation support system. However, the simulation was insufficient in that it did not consider delays in evacuation completion time due to road congestion caused by limited road width, which can be expected in an actual disaster. In our current research, as an improvement to make the simulation more realistic for further developing our evacuation support system, we propose a simulation that considers road width. We examined how adoption of this simulation methodology would affect the simulation results by comparing simulations that did and did not consider road width. As a result, the proposed method showed a lower percentage of completed evacuations than the previous simulation, especially for children and the elderly. The reason for this result is that when evacuations begin simultaneously, crowding occurs as people all rush to evacuation centers. This congestion on the roads near evacuation centers prevents people from completing their evacuations. Such a phenomenon is expected to occur not only in simulations but also on actual roads. For this reason, our proposed simulation method that considers road width is a more realistic simulation. Since evacuations by car in the event of a disaster also occur, a future issue is to simulate car congestion as well. Furthermore, simulating what would happen if a disaster made a road impassable is also important.
-
Proposal and evaluation for color constancy CAPTCHA Reviewed
Usuzaki S., Aburada K., Yamaba H., Katayama T., Mukunoki M., Park M., Okazaki N.
Artificial Life and Robotics 26 ( 3 ) 291 - 296 2021.8
Publishing type：Research paper (scientific journal) Publisher：Artificial Life and Robotics
CAPTCHAs are technologies that distinguish humans and bot to prevent illegal access. Unfortunately, current CAPTCHAs, even the latest Google reCAPTCHA, have already broken with high accuracy. Although the devices, including emphasizing the distortion of the text and adding noise to the image, improve the machine resistance, they may decrease the accessibility of the web page. The purpose of this study is to propose a new CAPTCHA that can decrease the machine resistance while keeping usability. To achieve this purpose, we focused on color constancy. Color constancy is a human’s characteristic that enables humans to recognize the original color of the object by ignoring the effects of illumination light. Color constancy has not been fully reproduced by the program yet. We proposed color constancy CAPTCHA that the user is required to answer an original color of the object in a specified area on the CAPTCHA image with a color filter. In this paper, we created a prototype of CAPTCHA, applied two kinds of color filters, and then evaluated each case for the human success rate, machine success rate, and usability.
-
Proposal of Jawi CAPTCHA Using Digraphia Feature of the Malay Language Reviewed
Yamaba H., Sohaimi A.S.A.B.A., Usuzaki S., Aburada K., Mukunoki M., Park M., Okazaki N.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 12835 LNCS 119 - 133 2021
Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
This paper proposes a new text-based CAPTCHA using Jawi script and Latin script, which are both used in the Malay language writing system. Many web sites have adopted CAPTCHA to prevent bots and other automated programs from malicious activities such as posting comment spam. Text-based CAPTCHA is the most common and earliest CAPTCHA. But as optical character recognition (OCR) technology has improved, the intensity of distortions that must be applied to a CAPTCHA for it to remain unrecognizable by OCR has increased. This has reached a point where humans are having difficulty recognizing CAPTCHA text. The idea of the proposed CAPTCHA is to generate two identical character strings, one written in Latin script and the other in Jawi script. Because some of the strings characters are hidden by obstacles, users need to combine both strings to solve this CAPTCHA. This idea uses the fact that most Jawi characters have one-to-one correspondence with Latin characters. A series of experiments was carried out to evaluated the performance of the proposed CAPTCHA. First, a computer program was developed with various software languages for the usability evaluation. The results showed that the average time to solve the CAPTCHA and the accuracy rates were acceptable compared with the indices reported in existing research. Next, two OCR programs were applied to the Jawi CAPTCHA, and it was demonstrated that they could not read the partially hidden Latin and Jawi strings. Lastly, we discussed the effectiveness of the proposed CAPTCHA by relating to the priming effect.
-
Blockchain-based malware detection method using shared signatures of suspected malware files Reviewed
Fuji R., Usuzaki S., Aburada K., Yamaba H., Katayama T., Park M., Shiratori N., Okazaki N.
Advances in Intelligent Systems and Computing 1036 305 - 316 2020
Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
Although rapid malware detection is very important, the detection is difficult due to the increase of new malware. In recent years, blockchain technology has attracted the attention of many people due to its four main characteristics of decentralization, persistency, anonymity, and auditability. In this paper, we propose a blockchain-based malware detection method that uses shared signatures of suspected malware files. The proposed method can share the signatures of suspected files between users, allowing them to rapidly respond to increasing malware threats. Further, it can improve the malware detection by utilizing signatures on the blockchain. In the evaluation experiment, we perform a more real simulation compared with our previous work to evaluate the detection accuracy. Compared with heuristic methods or behavior-based methods only, the proposed system which uses these methods plus signature-based method using shared signatures on the blockchain improved the false negative rate and the false positive rate.
-
On a user authentication method to realise an authentication system using s-EMG Reviewed
Yamaba H., Usuzaki S., Takatsuka K., Aburada K., Katayama T., Park M., Okazaki N.
International Journal of Grid and Utility Computing 11 ( 5 ) 725 - 734 2020
Publishing type：Research paper (scientific journal) Publisher：International Journal of Grid and Utility Computing
To prevent shoulder-surfing attacks, we proposed a user authentication method using surface electromyogram (s-EMG) signals, which can be used to identify who generated the signals and which gestures were made. Our method uses a technique called 'pass-gesture', which refers to a series of hand gestures, to achieve s-EMG-based authentication. However, it is necessary to introduce computer programs that can recognise gestures from the s-EMG signals. In this paper, we propose two methods that can be used to compare s-EMG signals and determine whether they were made by the same gesture. One uses support vector machines (SVMs), and the other uses dynamic time warping. We also introduced an appropriate method for selecting the validation data used to train SVMs using correlation coefficients and cross-correlation functions. A series of experiments was carried out to confirm the performance of those proposed methods, and the effectiveness of the two methods was confirmed.
MISC 【 display / non-display 】
-
Usuzaki S., Aburada K., Yamaba H., Katayama T., Mukunoki M., Park M., Okazaki N.
Artificial Life and Robotics 27 ( 1 ) 2022.2
Publishing type：Rapid communication, short report, research note, etc. (scientific journal) Publisher：Artificial Life and Robotics
In the original publication of the article, on Figs. 4 and 5, “CNN” should read as “MLP”. In addition, under the section “4.2 Human and machine success rate”, on the paragraph “For evaluation of the machine success …”, in the following sentence “In the attack experiment, we applied the famous color …” CNN (Convolutional Neural Network) should be corrected as MLP (Multilayer perceptron). The correct sentence should read as “In the attack experiment, we applied the famous color constancy algorithm, Gray-World, Max-RGB, Gray-Edge [11], Second Derivative Gray-Edge, and MLP (Multilayer perceptron) to images that were saved in the experiment for the human success rate
Grant-in-Aid for Scientific Research 【 display / non-display 】
-
透過的反射による人間の映り込みを高精度に検出するSNS投稿画像解析手法の開発
Grant number：23H05387 2023.04 - 2024.03
独立行政法人日本学術振興会 科学研究費補助金 奨励研究
Authorship：Principal investigator