Accelerometers built into devices, such as smartphones and smartwatches, are used for applications including activity tracking and health management. A method has been proposed that intentionally modifies sensor output by irradiating sound waves, and research exists that applies this technique to embed arbitrary bit sequences into capacitive accelerometers. This paper proposes a method for real-time analysis and identification of bit sequences embedded in acceleration data. The proposed method enables contactless control of smartphones through the use of sound wave irradiation. In the evaluation experiment, the success rate of correctly launching the appropriate application for 4 to 8-bit sound wave patterns was measured. While Devices 1 and 3 achieved 100% accuracy under all conditions, Device 2 showed a tendency for the success rate to decrease as the number of bits increased.

As it becomes easier to obtain various data from wearable devices, it is known that biometric and behavioral information must be handled with care. On the other hand, data on the surrounding environment, such as outside temperature, is seen as having a weak relationship with the wearer, and data handling is considered to be a chore. We believe that even data with weak relationships have the potential to infer information about the wearer if a large amount of data is acquired. In this paper, we verify whether it is possible to estimate the wearer's location from time series data of outside air temperature using only the temperature sensor. We calculated the average absolute error between the temperature data from the wearable device and the same time-series data obtained from the Japan Meteorological Agency, and we evaluated the wearer's position estimation. It was found that the location where the temperature was sampled appeared at the top of the estimation ranking, and that cities near the sampling location were estimated to be at the high ranking. It was also found that the number of data to be used can be a factor that increases the estimation ranking.

Many hands-free input methods using ear accessories have been proposed. Although most of previous studies use canal type earphones, we focus on the following two points. 1) A method for using the hands-free input function with ear accessories that are not canal type. 2) A method for using the same hands-free input function across multiple ear accessories. Then, using an infrared distance sensor attached to the ear accessory, we propose a method for recognizing the gesture of the user’s facial expression. Based on the change in distance between the infrared distance sensor attached to the ear accessory and the skin, the proposed method detects skin movement around the ear, which differs for each facial expression gesture. We created a prototype system for the root of the ear, earlobe, and tragus ear accessories. The evaluation result for nine gestures and five subjects showed that F-value was 0.94 or more for one device alone, and the F-value was 0.97 or more for the pattern combining multiple devices.

Sensing technology for eyeball movement (i.e., gaze movement) has enabled various applications, e.g., hands-free input interfaces and provided information that helps us understand human beings in various fields. However, the existing method, which senses eyeball movement constantly using a wearable device has limitations that should be addressed, e.g., cost. Therefore, we propose a method to recognize eyeball movements using eyewear equipped with multiple infrared distance sensors. The proposed method recognizes eyeball movement by sensing time-series data of eyelid skin movement, which accompanies eyeball movement, using an infrared distance sensor inside the eyewear. Evaluation results of five subjects demonstrate that the proposed method can recognize five types of movement with an F-value of 1.0 and 20 types of movement with an F-value of 0.94. Our study showed the feasibility of the proposed method for gaze input interface and for inexpensive technology that constantly senses the eyeball movement.

Research on hands-free input methods has been actively conducted. However, most of the previous methods are difficult to use at any time in daily life due to using speech sounds or body movements. In this study, in order to realize a hands-free input method based on nasal breath using wearable devices, we propose a method for recognizing nasal breath gestures, using piezoelectric elements placed on the nosepiece of a glasses-type device. In the proposed method, nasal vibrations generated by nasal breath are acquired as sound data from the devices. Next, the breath pattern is recognized based on the factors of breath count, time interval, and intensity. We implemented a prototype system for initial evaluation. The evaluation results for eight subjects showed that the proposed method can recognize eight types of nasal breath gestures with 0.89% of F value. Our study provides the first wearable sensing technology that uses nasal breathing for hands-free input.

Various types of helmets exist, including industrial protective helmets, motorcycle helmets, sports helmets, and military/police helmets. By identifying individuals wearing a helmet, their name, affiliation, and qualification can be presented on a display mounted on the helmet, and sensor data collected through the helmet, such as acceleration, video, and eye-tracking data, can be labeled with the user's ID. In this paper, we propose a user identification method based on head shape using a helmet equipped with 32 pressure sensors. We implemented a prototype helmet device and collected data from nine subjects, resulting in 100% accuracy for user identification and an average equal error rate of 0.076 for user authentication.

With the increasing use of wearable devices equipped with various sensors, human activities, biometric information, and surrounding situations can be obtained via sensor data regardless of time and place. When position-free wearable devices are attached to an arbitrary part of the body, the attached position should be identified because the application process changes relative to the position. For systems that use multiple wearable devices to capture body-wide movement, estimating the attached position of the devices is meaningful. Most conventional studies estimate the loading position of the sensor using accelerometer and gyroscope data; therefore, users must perform specific motions so that each sensor produces values unique to the given position. We propose a method that estimates the load position of wearable devices without forcing the wearer to perform specific actions. The proposed method estimates the time difference between a heartbeat obtained by an electrocardiogram and a pulse wave obtained using a pulse sensor and classifies the sensor position from the estimated time difference. We assume that pulse sensor is embedded in the wearable devices to be attached to the user. From the results of an evaluation experiment with five subjects, an average F-measure of 0.805 was achieved over 15 body parts. The left ear and the right finger achieved an F-measure of 0.9+ when the proposed system uses data of approximately 20 seconds as an input.

Along with the downsizing of computers and sensors, user situation and state can be recognized by sensing, signal processing, and machine learning technologies in many places. In home environments, it is reported that the number of drowned people in a bathtub has increased to 4,886 in 2004, which is 1.7 times from 2004. In order to prevent the occurrence of such a fatal accident, it is important to grasp the state of people while bathing. In this study, we propose a method to identify people in the bathtub from changes in water level during bathing with a water pressure sensor installed at the bottom of the bathtub. In the evaluation experiments, the average identification accuracy for four subjects is as low as 65% in the comparison method using only the change in the water level before and after entering the bathtub. The proposed methods uses DTW distance of the waveform of entering water section and exiting water section and its average accuracy achieved 95%.

Along with the progress of miniaturization and energy saving technologies of sensors, biological information in our daily life can be monitored by installing the sensors to a lavatory bowl. Lavatory is usually shared among several people, therefore biological information need to be identified. Using camera, microphone, or scales is not appropriate considering privacy in a lavatory. In this paper, we focus on the difference in the way of pulling a toilet paper roll and propose a system that identifies individuals based on features of rotation of a toilet paper roll with a gyroscope. The evaluation results confirmed that 85.8% accuracy was achieved for a five-people group in a laboratory environment.

Recently, bicycling as a sport has attracted a great deal of attention. Previous research on bicycles suggests that pedaling at high frequency at a constant speed is most effective. However, it is hard for beginners to acquire such pedaling skills since expert cyclists develop the skills through long-term training. We propose a bicycle pedaling training system using auditory feedback. The system generates feedback sound every time a pedal crank turns a quarter rotation. Users can keep the pedaling speed constant by synchronizing pedaling with the feedback sound with background music whose tempo is constant. We conducted an experiment with eleven subjects for four weeks and confirmed that the variances of pedaling speed for the subjects trained with the proposed system decreased significantly compared with those of the conventional method.

There have been several studies on object detection and activity recognition on a table conducted thus far. Most of these studies use image processing with cameras or a specially configured table with electrodes and an RFID reader. In private homes, methods using cameras are not preferable since cameras might invade the privacy of inhabitants and give them the impression of being monitored. In addition, it is difficult to apply the specially configured system to off-the-shelf tables. In this work, we propose a system that recognizes activities conducted on a table and identifies which user conducted the activities with load cells only. The proposed system uses four load cells installed on the four corners of the table or under the four legs of the table. User privacy is protected because only the data on actions through the load cells is obtained. Load cells are easily installed on off-the-shelf tables with four legs and installing our system does not change the appearance of the table. The results of experiments using a table we manufactured revealed that the weight error was 38 g, the position error was 6.8 cm, the average recall of recognition for four activities was 0.96, and the average recalls of user identification were 0.65 for ten users and 0.89 for four users.

In this paper, we propose a data forwarding method for efficient data gathering in wireless sensor networks with multiple mobile sinks which freely move in the target region. In our proposed method, each sensor node forwards data packets based on information on connectivity (the status of connection with mobile sinks). As the way to share information on connectivity among neighboring nodes, we integrate push-based and pull-based mechanisms we have proposed in [5]. The integrated mechanism utilizes advertisement from the nodes with good connectivity together with request from the nodes which hold data to transmit. By the complementary use of these mechanisms, our proposed method reduces the traffic for sharing information while realizing an efficient data forwarding.

An accelerometer is installed in most current mobile phones, such as iPhones, Android-powered devices, and video game controllers for the Wii or PS3, which enables easy and intuitive operations. Therefore, many gesture-based user interfaces that use accelerometers are expected to appear in the future. Gesture recognition systems with an accelerometer generally have to be models constructed with a user's gesture data before use, and they need to recognize any unknown gestures by comparing them with an output of the recognition result and feedback delays since the recognition process generally starts after the gesture has finished, which may cause users to retry gestures and thus degrade the interface usability. We propose an early stages gesture recognition method that sequentially calculates the distance between the input and training data, and outputs recognition results only when one output candidate has a stronger likelihood than the others. Gestures are recognized in the early stages of a given motion without deteriorating the level of accuracy, which improves the interface usability. Our evaluation results indicated that the recognition accuracy approached 1.00 and the recognition results were output 1,000 msec on average before a gesture had finished.

Authentication methods with password, PIN, face recognition, and fingerprint identification have widely been used, however these methods have problems of difficulty in one-handed operation, vulnerability to shoulder hacking, and illegal access using fingerprint with super glue or facial portrait. Prom viewpoint of usability and safety, strong and uncomplicated method is required. We propose a user authentication method based on grip gestures using pressure sensors mounted on the lateral and back of a mobile phone. Grip gesture is an operation of grasping a mobile phone, which is assumed to be done instead of conventional unlock procedure. Grip gesture can be performed with one hand. Moreover, it is hard to imitate grip gestures since finger movements and grip force during a grip gesture are hardly seen by the others. We experimentally investigated the feature values of grip force and evaluated our proposed method from viewpoint of error rate.

In various environments, such as mobile and wearable computing, compact I/O devices are desirable from the viewpoint of portability. Now, many users are accustomed to input with a keyboard, however, there is a limitation of miniaturization because it degrades the performance of key touch. Therefore, in this paper, we propose a method to miniaturize a keyboard by excluding the half of it. In using the proposed method, one hand hits keys as usual, and the other hand hits the place outside the keyboard as if the user types with both hands. The user can input words with only one hand because the proposed system estimates the input word using keying interval, which appears also when the user inputs with both hands. From the results of user study, we confirmed that the user can input with only one hand and that it does not decrease input speed drastically.

Mobile phones and video game controllers using gesture recognition technologies enable easy and intuitive operations, such as scrolling a browser and drawing objects. However, usually only one of each kind of sensor is installed in a device, and the effect of multiple homogeneous sensors on recognition accuracy has not been investigated. Moreover, the effect of the differences in the motion of a gesture has not been examined. We have investigated the use of a test mobile device with nine accelerometers and nine gyroscopes. We have captured the data for 27 kinds of gestures for a mobile tablet. We experimentally investigated the effects on recognition accuracy of changing the number and positions of the sensors and of the number and kinds of gestures. The results showed that the use of multiple homogeneous sensors has zero or negligible effect on recognition accuracy, but that using an accelerometer along with a gyroscope improves recognition accuracy. They also showed that some gestures were not consistent among test subjects and interdependent, so selecting specific gestures to use can improve recognition accuracy.

Human activity recognition through the wearable sensor will enable a next-generation human-oriented ubiquitous computing. However, most of research on human activity recognition so far is based on small number of subjects, and non-public data. To overcome the situation, we have gathered 4897 accelerometer data with 116 subjects and compose them as HASC2011corpus. In the field of pattern recognition, it is very important to evaluate and to improve the recognition methods by using the same dataset as a common ground. We make the HASC2011corpus into public for the research community to use it as a common ground of the Human Activity Recognition. We also show several facts and results of obtained from the corpus.

There are many kinds of events where participants converge and move around freely looking at points of interest. Since in these event spaces, the participants go where they want, some problems may arise for the event manager such as people staying in one area too long or congestion that occurs at specific attractions. Therefore, we propose here a new navigation system that has a route planning algorithm to satisfy the objectives of the event manager. In an actual test of our system, we found that the participants' actions were in line with the objectives of the event manager. Moreover, we also found that our system performs better by using wearable computing technologies because detailed information on participants is acquired.