Shigeyuki Hirai

Bathcratch: touch and sound-based DJ controller implemented on a bathtub

Bathcratch is a music entertainment system that converts a bathtub into a DJ controller, allowing an average person in a bathtub to play scratching music. The system detects the squeaks made by rubbing the bathtub and associates them with several preset scratching phrases. In addition, capacitive touch sensors embedded in the tub allow the selection of scratching phrases and background rhythm tracks. Here, we provide a system overview and explain the design, user interface, music controller implementation of this system along with the feedback received for it during a public exhibition.

Enabling Interactive Bathroom Entertainment Using Embedded Touch Sensors in the Bathtub

We propose a new entertaining bathroom environment with applications controlled via capacitive touch sensors embedded in the bathtub. The bathtub touch sensor system, called TubTouch, provides a new touch user interface near to the edge of the bathtub for persons who are bathing. TubTouch can be used to control both existing bathroom equipment, such as water heaters, jacuzzis, TVs, audio, and lighting, and a variety of new applications. In this paper, we give an overview of the TubTouch system and discuss its entertainment applications used in daily life.

Entertainment applications for tapping on a bathtub edge using embedded acoustic sensors

In past studies of smart environments and their applications in a house, we have developed bathtub systems that convert a normal tub into a user interface (UI) by the use of embedded sensors. We have developed another proposed UI for tapping, rapping, and knocking on a bathtub edge and some potential entertainment applications. In this paper, this bathtub system, which uses embedded piezoelectric acoustic sensors, is described, together with its entertainment applications in a bathroom.

Interaction Technique Using Acoustic Sensing for Different Squeak Sounds Caused by Number of Rubbing Fingers

Various studies have been conducted for developing interaction techniques in a smart house. Some of our previous studies [1, 2] focused on bathrooms and we converted an existing normal bathtub system into a user interface by using embedded sensors. A system called Bathcratch [2] detects squeak sounds by rubbing on a bathtub edge. To generate squeaks, it requires some conditions to cause the Stick-slip phenomenon. A bathtub has a smooth surface, and water can cause the phenomenon with human skins. Bathcratch uses it as an interaction technique to play DJ-scratching. Here, we extended the interaction technique using squeaks to recognize rubbing states, rubbing events including sequence, and the difference between squeaks caused by the number of fingers. This can be used in various wet environments including kitchen, washbowls in a restroom, swimming pool, and spa. This paper describes the method and its performance for identifying the number of rubbing fingers by using frequency analysis. In addition, we illustrate some smart home applications by using the proposed technique.

Multi-touch wall display system using multiple laser range scanners

In this work we present a multi-touch wall display system that equips easily and considers occlusions problem using multiple range scanners. Our system is implemented on an existing large display embedded in a wall with two laser range scanners. Each scanner detects touches events including positions and/or areas. Some touch positions causes relative occlusions due to a range scanner. We reduce the problem by using multiple scanners. If the touch events from different scanners are the same, these are combined into one touch event. Detected multi-touch events are sent to the network as TUIO events. This system is simple and adaptable to various existing displays including front projection screens, and to various TUIO applications.

RapTapBath: User Interface System by Tapping on a Bathtub Edge Utilizing Embedded Acoustic Sensors

We present RapTapBath: a human-computer interface system that converts an existing bathtub into a controller that recognizes hand-tapped tones and patterns on a bathtubs edge. This system utilizes embedded piezoelectric sensors in the bathtubs edge to analyze acoustic signals of tapped sounds, and projects a menu on the tubs edge using a projector installed above the tub. Tap locations are detected by measuring differences of signal propagation times using the piezoelectric sensors. Tap tones are identified by spectrum patterns using non-negative matrix factorization. Tap patterns are detected via tapping rates, whose probability calculations are measured over a fixed duration. This paper describes the tapping user interfaces (UI) events and their specific detection methods and signal processing techniques. We also present an experimental performance evaluation. We also propose effective applications for spending bath time using this system. Finally, we describe the limitations of current tap UI events and the implications of interaction designs of RapTapBath.

BathDrum2: Percussion Instruments on a Bathtub Edge with Low-Latency Tap Tone Identification

In order to create a smart bathroom with entertainment features, we propose a playable system that can identify several tap tones on the top of a bathtub edge from differences in hand posture. BathDrum2 is a specific implementation of this system; it utilizes nonnegative matrix factorization (NMF), a kind of machine learning technique, for low-latency tap tone identification and for playing assigned percussion sounds. This paper describes the system design and tap tone identifier with its signal processing and NMF. The performance evaluation of the current BathDrum2 implementation is also described.

Editor's Message to Special Issue on Extensions and Advances in Music Information Processing

There has been the great progress and expansion both in depth and width in music information processing research. Recent sophistication of acoustic signal processing and machine learning techniques, realization of large-scale application systems, such as Apple Music and Google Play Music, which leverage information on the Internet have contributed to the progress of such research developments. On the other hand, research in music creation has also become popular, due to the spread of new musical instruments/systems with physical interfaces developed as a part of the current digital fabrication movement, and also from new ideas generated by utilizing ideas in other fields of research. This special issue of music information processing is the first for the IPSJ journal which includes audio/video data. The audio/video data have been handled as a part of the submitted papers, i.e., were subject to peer-reviewing along with the papers, and are archived with the accepted papers of this edition. Considering the characteristics of music, it is very important to archive and convey not just the paper, but also specific contents of the research results accurately and clearly, which was impossible with only conventional paper media. This special issue was planned to increase the value of academic reference materials with this aim. While this objective may not be applicable to all research in the computer science area, we feel that this trial is essential for IPSJ to provide academic information services in the multimedia era. Out of 29 submissions, seven papers were accepted (i.e. acceptance rate 24%) by the editorial committee of this special issue that was organized from the committee of the Special Interest Group of Music Information Processing. Two papers were submitted with media data and both were accepted for publication in this issue. We would like to thank all the authors for their contributions, and reviewers for short-term peer-reviewing. We also would like to thank to the editorial committee and secretaries of IPSJ.