- Tactile and Sensory Interactions
- AI in Service Interactions
- Hand Gesture Recognition Systems
- Speech and Audio Processing
- Speech and dialogue systems
- Digital Media Forensic Detection
- User Authentication and Security Systems
- Topic Modeling
- Interactive and Immersive Displays
- Data Visualization and Analytics
- Digital Communication and Language
- Orbital Angular Momentum in Optics
- Crime, Illicit Activities, and Governance
- Media, Communication, and Education
- Robotics and Sensor-Based Localization
- Gaze Tracking and Assistive Technology
- Teleoperation and Haptic Systems
- History and Developments in Astronomy
- Usability and User Interface Design
- Subtitles and Audiovisual Media
- Internet Traffic Analysis and Secure E-voting
- Indoor and Outdoor Localization Technologies
- Recommender Systems and Techniques
- Speech Recognition and Synthesis
- Smart Grid Security and Resilience
Waseda University
2018-2024
National Institute of Advanced Industrial Science and Technology
2023-2024
University of Tsukuba
2021-2023
University of Crete
2021
FORTH Institute of Computer Science
2021
The Open University of Japan
2021
Tokyo Institute of Technology
2020
National Institute of Information and Communications Technology
2019
Gunma University
1976
We propose a novel attack, called an “Audio Hotspot Attack,” which performs inaudible malicious voice command by targeting assistance systems, e.g., smart speakers or in-car navigation systems. The key idea of the approach is to leverage directional sound beams generated from parametric loudspeakers, emit amplitude-modulated ultrasounds that will be self-demodulated in air. Our work goes beyond previous studies attack following three aspects: (1) can succeed on long distance (3.5 meters...
In this study, we conduct the first study on analysis of voice assistant (VA) apps. We collect metadata VA apps from app directory and analyze them. Next, call by corresponding commands examine how they identify users analyzing responses found that roughly half performed user identification some means. also several aim to acquire personal information such as birth date, age, or blood type through conversations. As data will be stored in cloud, need have a mechanism ensure an end-user can...
We propose a novel attack named "Audio Hotspot Attack'', which performs an inaudible malicious voice command attack, targeting assistance systems, e.g., smart speakers or in-car navigation systems. This leverages directional sound beams generated from parametric loudspeakers, emit AM-modulated ultrasounds that will be self-demodulated in the air. It can succeed on long distance (2--4 meters small room and 10+ meter hallway). To evaluate feasibility of we performed extensive in-lab...
Deaf and hard of hearing (DHH) people have limited access to auditory input, so they mainly receive visual information during online meetings. In recent years, the usability a system that visualizes ongoing topic in conference has been confirmed, but it not verified remote includes DHH people. One possible reason is dispersion occurs when there are multiple sources information. this study, we introduce "Word Cloud for Meeting," generates separate word cloud each participant displays...
This study introduces PinchKey, a highly accurate and user-friendly behavioral biometric authentication method for VR users. Authentication is performed using the natural intuitive 'pinch' gesture used when operating devices. Since this two-finger action devices, we expect that use of pinch will provide enhanced usability compared to traditional device methods such as pattern lock or PIN entry. The system analyzes three-dimensional trajectory user's finger movements during applies supervised...
When playing musical instruments, deaf and hard-of-hearing (DHH) people typically sense their music from the vibrations transmitted by instruments or movements of bodies while performing. Sensory substitution devices now exist that convert sounds into light to support DHH people's activities. However, these require specialized hardware, marketing profiles assume standard are available. Hence, a significant gap remains between performance enjoyment. To address this issue, study identifies end...
Social interaction begins with the other person's attention, but it is difficult for a d/Deaf or hard-of-hearing (DHH) person to notice initial conversation cues. Wearable visual devices have been proposed previously. However, these are cumbersome wear must stay within DHH vision. In this study, we SHITARA, novel accessibility method air vortex rings that provides non-contact haptic cue person. We developed proof-of-concept device and determined ring's accuracy, noticeability comfortability...
Smartphone applications that allow users to enjoy playing musical instruments have emerged, opening up numerous related opportunities. However, it is difficult for deaf and hard of hearing (DHH) people use these apps because limited access auditory information. When using real instruments, DHH typically feel the music from vibrations transmitted by or movements body, which not possible when with apps. We introduce "smartphone drum,'' a smartphone application presents drum-like vibrotactile...
This work developed a new security framework named Cyber-Physical Firewall (CPFW), which provides generic and flexible access control mechanism for regulating the malicious analog signals that target cyber-physical system (CPS) devices. enables defeat of various attacks make use against CPS devices; e.g., stealth voice command injection attack using ultrasonic waves or adversarial examples, to crash drones in flight sound waves.
Banks offering Online Banking services need to detect and prevent unauthorized electronic funds transfers reduce financial crime risk. They monitor online banking transaction histories use their own methods fund transfers. However, by criminals have not been eliminated. The average rate of false positives in the monitoring systems installed Japanese banks is up 99%, indicating that are fully functional. Furthermore, personnel responsible for fraud detection must manually check a large number...
The rate of absorption with chemical reaction, the interfacial area and their dependence on process variables have been investigated.The effective per unit bed volume increases increasing gas velocity up to 200 cm/sec (gas holdup-0.6), but decreases at a high velocity. seems be independent such as free opening supporting grid, static packing height, column diameter, solid holdup size density.
A voice assistant (VA) is a platform that provides users with wide range of services via interaction application using verbal commands. Since the VA deployed in cloud, its behavior not transparent to user, which raises privacy concerns. In this study, we developed framework called ChatterBox, attempts analyze applications extensive continuous interaction, understand their behavior. ChatterBox capable parsing and generating dialogues by utilizing natural language processing approach. It can...