Papers - KATAYAMA Tetsuro
-
LLMを用いたユニットテスト生成における実行時エラー抑制のためのプロンプト手法の提案 Reviewed
山田泉樹, 片山徹郎, 高橋 寿一
ソフトウェアテストシンポジウム2026東京(JaSST'26 Tokyo)論文集 7p 2026.3
Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
-
Takahashi T., Takahashi N., Katayama T., Kita Y.
Proceedings of International Conference on Artificial Life and Robotics 298 - 302 2026.2
Authorship：Corresponding author Publishing type：Research paper (scientific journal) Publisher：Proceedings of International Conference on Artificial Life and Robotics
Writing test cases from the formal specification description VDM++, which is a method for eliminating ambiguity in software specification, is time-consuming and labor-intensive. Therefore, our laboratory has developed BWDM, a tool that automatically generates test cases from VDM++ specification. However, the existing BWDM has a problem in that it cannot generate test cases from VDM++ specification having multiple classes. Therefore, in order to improve the usefulness of BWDM, this study extends BWDM to solve the above problem. Consequently, it has confirmed that the extended BWDM can reduce the test case generation time compared to manual test case generation for VDM++ specification having multiple classes.
-
Takuma C., Takahashi N., Katayama T., Kita Y.
Proceedings of International Conference on Artificial Life and Robotics 303 - 307 2026.2
Authorship：Corresponding author Publishing type：Research paper (scientific journal) Publisher：Proceedings of International Conference on Artificial Life and Robotics
We have developed ASLA (Automatic Segmentation and Labeling tool using AI) to divide electronic documents into elements and generate labels for new applications. However, the existing ASLA has two problems: (1) low segmentation accuracy for document formats other than scientific papers, and (2) occasionally generating incorrect labels. This study generates a new document image dataset with three additional formats, generates a segmentation model, and extends ASLA to use a multimodal LLM (Gemini API) for label generation. We have confirmed through application examples that extended ASLA can divide the new document formats and generate labels for them. Finally, evaluation results have shown that the generated segmentation model achieved a score of at least 0.78 in mAP@[0.75] for each of the four format types in the test data. Additionally, it has confirmed that the multimodal LLM generates correct labels with higher accuracy than the existing ASLA method.
-
A Design Testing Method Using Formal Models Based on EPNAT and ESTG for Web Applications Reviewed
Amo T., Katayama T., Takagi T.
Proceedings of International Conference on Artificial Life and Robotics 294 - 297 2026.2
Publishing type：Research paper (scientific journal) Publisher：Proceedings of International Conference on Artificial Life and Robotics
This study shows a method to test designs of Web applications using formal models. In the method, designs of servers and clients of a Web application are defined as Extended Place/transition Net with Attributed Tokens (EPNAT) and Extended Screen Transition Graph (ESTG) models, respectively. The EPNAT model represents concurrent distributed behavior to provide services to the clients, and the ESTG model represents behavior to request and receive the services through a Web browser. A test data generator based on user models and test strategies interacts with the ESTG models, and the ESTG models interact with the EPNAT model. The EPNAT and ESTG models contain assertions that detect failures on the designs, that is, detect violations of constraints derived from specifications. The procedure, effectiveness, and future work are discussed in this paper.
-
Matsumoto Sho, Katayama Tetsuro, Takagi Tomohiko
Journal of Robotics, Networking and Artificial Life 11 ( 2 ) 175 - 179 2025.12
Language：English Publishing type：Research paper (scientific journal) Publisher：The Society of Artificial Life and Robotics
The collaborative behavior of multiple systems provides valuable functions and services to users. However, it is actualized by large and complex implementations, which frequently include serious failures. In this study, we propose an automated random simulation (ARS) technique for checking the abstract collaborative behavior of multiple systems at the design level. The abstract collaborative behavior is expected to be designed using an extended place/transition net with attributed tokens (EPNAT), and the checking is performed dynamically based on the design called "EPNAT model". The ARS technique consists of (1) an algorithm for model execution using random search with the evaluation of constraints including feasibility, and (2) a stopping criterion for model execution focusing on glue transitions. The ARS technique requires tool support; therefore, we developed a prototype tool. We ran the prototype tool with a trial model and three faulty models, and found its effectiveness and future challenges.
-
Conjugate Gradient Based Multiplicative Update Rules for Nonnegative Matrix Factorization Reviewed
Takahashi N., Usuzaki S., Katayama T., Yokomichi M., Aburada K., Okazaki N.
Proceedings 9th International Conference on Information Technology Incit 2025 639 - 644 2025.11
Publishing type：Research paper (scientific journal) Publisher：Proceedings 9th International Conference on Information Technology Incit 2025
In this paper, we propose a general solution for nonnegative matrix factorization (NMF), a technique widely used in the field of machine learning. NMF is a type of matrix factorization used for dimensionality reduction. By approximating a nonnegative matrix as the product of two nonnegative matrices, it is possible to extract frequent patterns in the data matrix into the basis matrix together with their corresponding weights. Since many types of information, such as images, spectrograms of acoustic signals, and text data, are represented by nonnegative values, NMF is a useful algorithm applicable across a wide range of fields. Standard approaches to solving NMF include algorithms based on multiplicative update rules. However, because the applicable algorithm depends on the problem structure, extending NMF to new problem settings can be challenging. Another approach is the gradient method, which is an additive iterative procedure for solving a nonlinear optimization problem. The update direction of the state is given by the gradient of the objective function. The gradient method is versatile. It can be applied to any optimization problem with an available gradient, regardless of the problem structure. However, when applying it to the NMF problem, it is necessary to maintain the nonnegativity of the matrices throughout the iterative computation. As a result, several issues must be addressed, including vanishing gradients, proper tuning of the learning rate, and the need for gradient clipping. In this paper, we propose a new algorithm based on a multiplicative update rule incorporating the gradient. A multiplier function consisting of the hyperbolic tangent function and the linear functions is used. This function ensures that the pair of factor matrices remain nonnegative throughout the entire computation process. As with general gradient methods, the proposed algorithm has the advantage of being widely applicable to NMF problems formulated in a differentiable form. Furthermore, the convergence speed is improved by utilizing the nonlinear conjugate gradient method. Numerical experiments show that the decomposed matrices have a property specified by the penalty terms and confirming the effectiveness of the proposed method.
-
Webページのレイアウト不具合を強調表示する視覚的回帰テストツールMixVRTのクロスブラウザテスト対応による実用性の向上 Reviewed
有留直希, 片山徹郎
ソフトウェアエンジニアリングシンポジウム2025論文集 209 - 215 2025.9
Authorship：Corresponding author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
-
電子フォーム作成を目的とした記入欄自動検出およびラベル割付ツールの開発
木村優哉, 片山徹郎, 猿谷吉行, 中武範彰
情報処理学会第204回マルチメディア通信と分散処理・第109回電子化知的財産・社会基盤合同研究会 2025-DPS-204 ( 9 ) 8p 2025.9
Authorship：Corresponding author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
-
Yamaba H., Komori N., Usuzaki S., Aburada K., Katayama T., Okazaki N.
Icce Taiwan 2025 12th IEEE International Conference on Consumer Electronics Taiwan Generative AI in Innovative Consumer Technology Proceedings 111 - 112 2025.6
Publishing type：Research paper (scientific journal) Publisher：Icce Taiwan 2025 12th IEEE International Conference on Consumer Electronics Taiwan Generative AI in Innovative Consumer Technology Proceedings
Mobile devices such as smartphones and tablets are now deeply integrated into daily life. To prevent shoulder surfing attacks, we propose a user authentication method for mobile devices based on surface electromyogram (s-EMG) signals. The method uses a “pass-gesture,” a sequence of hand gestures that can be changed like a password. Fingerspelling was adopted as the source of gesture candidates. Previous studies confirmed high accuracy in identifying gestures performed by the same subject. This paper focuses on evaluating subject specificity-specifically, confirming that a model trained on one user fails to recognize gestures performed by another. Experimental results show that the system does not generalize across users, indicating robustness against impersonation.
-
Aridome N., Katayama T., Kita Y., Yamaba H., Aburada K., Okazaki N.
Proceedings of International Conference on Artificial Life and Robotics 575 - 579 2025.2
Language：English Publishing type：Research paper (scientific journal) Publisher：Proceedings of International Conference on Artificial Life and Robotics
As a method for detecting layout defects in web pages, image-based visual regression testing is proposed. However, it has the problem that it takes time to detect unintended layout differences that are not based on HTML code changes. This paper proposes a prototype of MixVRT which is a tool to detect layout defects in web pages. It is a visual regression testing tool that highlights layout defects in web pages. From evaluation experiments, the time find to layout defects can be reduced.
-
Kakinoki K., Katayama T.
Lecture Notes in Electrical Engineering 1322 LNEE 375 - 384 2025.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Electrical Engineering
To extract important information from the document images, document layout analysis research has been carried out. Previous research analyzes document layouts only for specific document formats. This paper proposes a document layout analysis method that adopts pre-processing based on document image classification to analyze document layouts for various document formats. From the application examples, we have confirmed that the proposed method can correctly analyze document layout by using training models trained for each document format. Furthermore, to evaluate the usefulness of the proposed method, we have compared the F score of the proposed method for document layout analysis using a single training model covering various document formats and using a training model generated for each document format. As a result, we have confirmed that our proposed method is useful in document layout analysis.
-
Kimura Y., Katayama T., Kita Y., Yamaba H., Aburada K., Okazaki N.
Proceedings of International Conference on Artificial Life and Robotics 580 - 584 2025.2
Language：English Publishing type：Research paper (scientific journal) Publisher：Proceedings of International Conference on Artificial Life and Robotics
The digitalization of forms is being promoted. One of the effective ways to manage contents filled in fields is using electronic forms. Several tools have been developed to generate them automatically. However, when you use a paper form, the layout of the original form may change, and it takes time to generate electronic one because it is necessary to place fill-in fields on an electronic form by dragging with a mouse. This paper proposes a method for automatic fill-in fields detection and labels assignment to reduce time required to place fill-in fields without changing the layout. The proposed method can reduce the time to place fill-in fields.
-
Evacuation Support System for Tsunami Disasters that Considers Shelter Congestion Reviewed
Nishi H., Nabeyama K., Usuzaki S., Aburada K., Yamaba H., Katayama T., Park M., Okazaki N.
Lecture Notes in Electrical Engineering 1322 LNEE 51 - 62 2025.2
Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Electrical Engineering
When a tsunami occurs, an evacuation support system capable of providing information on evacuation routes and shelter locations would enable evacuations to be conducted more swiftly and safely. In previous research, a proposal was made for an evacuation support system using low-power, long-distance communication. However, a challenge arose when shelter locations reached full capacity, requiring information to be shared, which in turn necessitated redirection of evacuees to alternative shelters. Therefore, in this study, a method for sharing shelter congestion information at an earlier stage was developed. The method of sharing based on the remaining capacity of shelters increased evacuation completion by up to 29 evacuees in an evacuee population of 1000.
-
Extending BWDM to Support Various Types and Recursive Definitions in VDM++ Test Case Generation Reviewed
Takakura Shota, Katayama Tetsuro, Kita Yoshihiro, Yamaba Hisaaki, Aburada Kentaro, Okazaki Naonobu
Journal of Advances in Artificial Life Robotics 4 ( 3 ) 167 - 171 2025
Language：English Publishing type：Research paper (scientific journal) Publisher：株式会社 ALife Robotics
Generating test cases from VDM++ formal specifications, which help to eliminate ambiguities, is
both time-consuming and labor-intensive. To solve this problem, our laboratory has developed
BWDM, a tool for automatic test case generation from VDM++ specifications. However, the
original BWDM only supports integer types and cannot handle test cases for operations and
functions with recursive structures. To enhance BWDM's usefulness, this paper introduces
extensions to address these limitations. The results confirm that the extended BWDM can reduce
test case generation time compared to manual methods. -
画面遷移システムを対象としたVDM++仕様記述のための入力テンプレートと 変換ルールの提案 Reviewed
高倉翔太, 片山徹郎
ソフトウェアエンジニアリングシンポジウム2024論文集 182 - 189 2024.9
Authorship：Corresponding author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
-
iOSアプリ開発支援を目的としたソフトウェア可視化ツールRAGESSの試作 Reviewed
翁長春樹, 片山徹郎
ソフトウェアエンジニアリングシンポジウム2024論文集 92 - 101 2024.9
Authorship：Corresponding author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
-
Onaga Haruki, Katayama Tetsuro, Kita Yoshihiro, Yamaba Hisaaki, Aburada Kentaro, Okazaki Naonobu
Journal of Robotics, Networking and Artificial Life 10 ( 4 ) 336 - 341 2024.9
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ALife Robotics Corporation Ltd.
It is difficult for mobile application developers to understand the structure of large and complex mobile applications. To support iOS application development, we proposed SwiftDiagram: a visualization of the static structure of Swift source code, and demonstrated its usefulness. To further support them, this paper has implemented RAGESS(Real-time Automatic Generation of SwiftDiagram System), which is a software visualization tool. RAGESS performs static analysis on Swift source code and automatically generates the corresponding SwiftDiagram whenever the target project build succeeds.
-
ASLA: Automatic Segmentation and Labeling by Deep Learning for Document Pictures Reviewed
Kakinoki Kanta, Katayama Tetsuro, Kita Yoshihiro, Yamaba Hisaaki, Aburada Kentaro, Okazaki Naonobu
Journal of Robotics, Networking and Artificial Life 10 ( 4 ) 362 - 367 2024.9
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ALife Robotics Corporation Ltd.
In this paper, we propose ASLA, a segmentation and label generation system for document pictures. ASLA reduces the duration needed for separating document pictures into areas and label generation. By using the application example, we have verified that ASLA operates properly. We have evaluated the usefulness of ASLA in regard to time and accuracy. We have assessed the efficacy of the rule-based area correction method. As a result, we have verified that ASLA is useful.
-
Usability improvement in color constancy CAPTCHA Reviewed
Usuzaki Shotaro, Yihan Wang, Aburada Kentaro, Yamaba Hisaaki, Takatsuka Kayoko, Katayama Tetsuro, Park Mirang, Okazaki Naonobu
IEICE Communications Express 13 ( 8 ) 331 - 334 2024.8
Language：English Publishing type：Research paper (scientific journal) Publisher：一般社団法人 電子情報通信学会
CAPTCHAs have traditionally been used to deter bots, but ensuring protection against machine learning attacks has become difficult. To tackle this problem, we proposed a CAPTCHA based on color constancy, a human cognitive ability that is difficult for machines to reproduce. Although this method achieved high attack resistance while maintaining usability compared to existing CAPTCHAs, the response time increased due to the color selection operation. To address this, we changed the task to a click-type format while keeping the original concept. Our experimental results show that our method matches the response times of existing CAPTCHAs without compromising attack resistance.
-
A Genetic Algorithm-based Approach for Design-level Class Decomposition Reviewed
Priyambadha B., Takahashi N., Katayama T.
International Journal of Advanced Computer Science and Applications 15 ( 4 ) 461 - 468 2024.4
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Advanced Computer Science and Applications
Software is always changed to accommodate environmental changes to preserve its existence. While changes happen to the software, the internal structure tends to decline in quality. The refactoring process is worth running to preserve the internal structure of the software. The decomposition process is a suitable refactoring process for Blob smell in class. It tried to split up the class based on the context in order to arrange it based on each responsibility. The previous approach has been implemented but still leaves problems. The optimum arrangement of class cannot be achieved using the previous approach. The genetic algorithm provides the search mechanism to find the optimum state based on the criterion stated at the beginning of the process. This paper presents the use of genetic algorithms to solve the design-level class decomposition problem. The paper explained several points, including the conversion from class to the chromosome construct, the fitness function calculation, selection, crossover, and mutation. The results show that the use of a genetic algorithm was able to solve the previous problems. The genetic algorithm can solve the local optimum problem from the previous approach. The increment of the fitness function of the study case proves it.