We present a wearable, oscillating magnetic field-based proximity sensing system to monitor social distancing as suggested to prevent COVID 19 spread (being between 1.5 and 2.0m) apart. We evaluate the system both in controlled lab experiments and in a real life large hardware store setting. We demonstrate that, due physical properties of the magnetic field, the system is much more robust than current BT based sensing, in particular being nearly 100% correct when it comes to distinguishing between distances above and below the 2.0m threshold.

This paper investigates the possibility of using soft smart textiles over the hair regions to detect chewing activities under episodes of snacking in a simulated scenario with everyday activities. The planar pressure textile sensors are used to perform mechanomyography of the temporalis muscles in the form of a cap. 10 participants contributed 30 recording sessions with time periods between 30 and 60 minutes. A frequency analysis method is developed to detect moments of snacking events with continuous sliding windows on 1-second time granularity. Our approach results in a baseline 80% accuracy, over 85% after outlier removal, and above 90% accuracy for some of the participants.

Many human activities and states are related to the facial muscles’ actions: from the expression of emotions, stress, and non-verbal communication through health-related actions. such as coughing and sneezing to nutrition and drinking. In this work, we describe, in detail, the design and evaluation of a wearable system for facial muscle activity monitoring based on a re-configurable differential array of stethoscope-microphones. In our system, six stethoscopes are placed at locations that could easily be integrated into the frame of smart glasses. The paper describes the detailed hardware design and selection and adaptation of appropriate signal processing and machine learning methods. For the evaluation, we asked eight participants to imitate a set of facial actions, such as expressions of happiness, anger, surprise, sadness, upset, and disgust, and gestures, like kissing, winkling, sticking the tongue out, and taking a pill. An evaluation of a complete data set of 2640 events with 66% training and a 33% testing rate has been performed. Although we encountered high variability of the volunteers’ expressions, our approach shows a recall = 55%, precision = 56%, and f1-score of 54% for the user-independent scenario(9% chance-level). On a user-dependent basis, our worst result has an f1-score = 60% and best result with f1-score = 89%. Having a recall ≥60% for expressions like happiness, anger, kissing, sticking the tongue out, and neutral(Null-class).

We investigate how pressure-sensitive smart textiles, in the form of a headband, can detect changes in facial expressions that are indicative of emotions and cognitive activities. Specifically, we present the Expressure system that performs surface pressure mechanomyography on the forehead using an array of textile pressure sensors that is not dependent on specific placement or attachment to the skin. Our approach is evaluated in systematic psychological experiments. First, through a mimicking expression experiment with 20 participants, we demonstrate the system’s ability to detect well-defined facial expressions. We achieved accuracies of 0.824 to classify among three eyebrow movements (0.333 chance-level) and 0.381 among seven full-face expressions (0.143 chance-level). A second experiment was conducted with 20 participants to induce cognitive loads with N-back tasks. Statistical analysis has shown significant correlations between the Expressure features on a fine time granularity and the cognitive activity. The results have also shown significant correlations between the Expressure features and the N-back score. From the 10 most facially expressive participants, our approach can predict whether the N-back score is above or below the average with 0.767 accuracy.

There have been many studies in recent years using the Textile planar Pressure Mapping (TPM) technology for computer-human interactions and ubiquitous activity recognition. A TPM sensing system generates a time sequence of spatial pressure imagery. We propose a novel, comprehensive and unified feature set to evaluate TPM data from the space and time domain. The initial version of the TPM feature set presented in this paper includes 663 temporal features and 80 spatial features. We evaluated the feature set on 3 datasets from past studies in the scopes of ambient, smart object and wearable sensing. The TPM feature set has shown superior recognition accuracy compared with the ad-hoc algorithms from the corresponding studies. Furthermore, we have demonstrated the general approach to further reduce and optimise the feature calculation process for specific applications with neighbourhood component analysis.

In this work, we present and evaluate a concept for using an integrated environment sensor as a wearable spirometer. Unlike a standard spirometer that by design is fairly bulky, our device can be unobtrusively integrated into various configurations suitable for long-term use in everyday settings (open headset, regular face mask, and professional sports mask). The sensor measures the transient change in air pressure, humidity and temperature in front of wearers’ mouth and nostrils. We present our hardware design and signal analysis methods needed to extract breathing rate information. We compare the results with a standard spirometer. Moreover, a calibration between the BME280 sensor and the spirometer is performed, having both working in parallel. We show that our approach is able to distinguish between normal breaths and deep breaths, as well as to capture the period and magnitude of the breath cycles, with a wearable device that can be used in everyday scenarios, as well as sport activities. The classification accuracy is 96% in face mask settings and 82% in an open headset setting. We also show that the sensor is able to approximate air volume by comparing the sensor’s pressure channel to the spirometer’s flow rate results.

We present a pervasive sensing system in the form of a wireless smart blanket that unobtrusively monitors people’s posture on ergonomic design chairs by covering the back piece of the chair and measuring the pressure profile of the user’s back. The sensor system has 1024 sensitive points, covering 48-by-48 cm² area. With simple and efficient classification algorithm, we reach around 80% among 10 postures, including lordotic and kyphotic lumber spine on different degrees, and lean to the sides on different degrees. The web browser based user interface offers timely and reprogrammable intervention from the user’s posture history.

Convolutional Neural Networks (CNNs) have become the state-of-the-art in various computer vision tasks, but they are still premature for most sensor data, especially in pervasive and wearable computing. A major reason for this is the limited amount of annotated training data. In this paper, we propose the idea of leveraging the discriminative power of pre-trained deep CNNs on 2-dimensional sensor data by transforming the sensor modality to the visual domain. By three proposed strategies, 2D sensor output is converted into pressure distribution imageries. Then we utilize a pre-trained CNN for transfer learning on the converted imagery data. We evaluate our method on a gait dataset of floor surface pressure mapping. We obtain a classification accuracy of 87.66%, which outperforms the conventional machine learning methods by over 10%.

In this paper, we present an approach for person identification using morphing footsteps measured from a fabric-based pressure mapping sensor system. The flexible fabric sensor is 0.5 mm thin and operates under a 5 mm thick normal carpet; therefore, it can be easily implemented into modern smart living spaces. We extract features concerning single steps with the shifting of gravity center, maximum pressure point and overall pressed area, which are independent from shape details and inter-step relationships of the walking sequences. The system is evaluated with 13 participants wearing shoes and walking normally across the carpet. Overall 529 footsteps are recorded, and the resulting average identification accuracy is 76.9%. Our approach can also be used for further activity recognition with the same physical carpet sensors.

In this paper, we developed a fully textile sensing fabric for tactile touch sensing as the robot skin to detect humanrobot interactions. We defined seven gestures which are inspired by the interactions of typical people to people and pet scenarios. Through evaluation experiment with 5 participants, the machine learning algorithm can recognize the gesture with a 98% accuracy if the algorithm has encountered the person before, and 86.8% accuracy if the person is excluded in the training data.

Based on a series of projects with textile pressure mapping matrix (TPM) for ubiquitous and wearable activity recognition in various scenarios, we have accumulated the knowledge and experience to develop an open-access hardware and software framework, which enables a broader education and allows the scientific community to build their own TPM applications. The hardware framework includes all the necessary resources to manufacture the sensing equipment and instructions to build the fabric sensors for an up to 32× 32 TPM. The software framework ‘Textile-Sandbox’contains ready-to-use tools and modules that support both running experiments and data mining. The framework is evaluated with 10 master students working in 4 groups. 4 applications are developed from scratch and validated within only 40 hours. We present this framework and the evaluated applications in this paper.

We investigate a generic fabric material as basis for resistive pressure and bio-impedance sensors and apply the fabric in a shirt collar for swallowing spotting. A pilot study confirmed the signal performance of both sensor types.

We present a wearable textile sensor system for monitoring muscle activity, leveraging surface pressure changes between the skin and an elastic sport support band. The sensor is based on an 8×16 element fabric resistive pressure sensing matrix of 1cm spatial resolution, which can be read out with 50fps refresh rate. We evaluate the system by monitoring leg muscles during leg workouts in a gym out of the lab. The sensor covers the lower part of quadriceps of the user. The shape and movement of the two major muscles (vastus lateralis and medialis) are visible from the data during various exercises. The system registers the activity of the user for every second, including which machine he/she is using, walking, relaxing and adjusting the machines; it also counts the repetitions from each set and evaluate the force consistency which is related to the workout quality. 6 people participated in the experiment of overall 24 leg workout sessions. Each session includes cross-trainer warm-up and cool-down, 3 different leg machines, 4 sets on each machine. Plus relaxing, adjusting machines, and walking, we perform activity recognition and quality evaluation through 2-dimensional mapping and the time sequence of the average force. We have reached 81.7% average recognition accuracy on a 2s sliding window basis, 93.3% on an event basis, and 85.6% spotting F1-score. We further demonstrate how to evaluate the workout quality through counting, force pattern variation and consistency.

We present a novel sensor system for the support of nutrition monitoring. The system is based on smart table cloth equipped with a fine grained pressure textile matrix and a weight sensitive tablet. Unlike many other nutrition monitoring approaches, our system is unobtrusive, non privacy invasive and easily deployable in every day life. It allows the spotting and recognition of food intake related actions, such as cutting, scooping, stirring, etc., the identification of the plate/container on which the action is executed, and the tracking of the weight change in the containers. In other words, we can determine how many pieces are cut on the main dish plate, how many are taken from the side dish, how many sips are taken from the drink, how fast the food is being consumed and how much weight is taken overall. In addition, the distinction between different eating actions, such as cutting, scooping, poking, provides clues to the type of food taken and the way the meal is consumed. We have evaluated our system on 40 meals (5 subjects) in a real life living environment: for seven eating related actions (cutting, scooping, stirring, etc.), resulting in above 90% average recognition rate for person dependent cases, and spotting each action out of continuous data streams (average F1 score 87%).

In this paper we use ubiquitous smart-chairs to evaluate live presentations. We validate the hypothesis that the audiences' activities can be recognized with pressure sensors under chairs' legs (74.6% accuracy rate from 8 typical activities in 8 live presentations, each with 6 chairs seated), and certain activities (eg talk, take notes) are linked to the audiences' subjective attitudes and evaluation of the presentation, moreover, the combinations of interactive activities like Talk-Take Note, Nod-Laugh and Take Note-Laugh are well linked to a positive rating of the presentation.

We demonstrate through a pressure sensor matrix, that weight distribution on feet is influenced by body posture. A small cheap carpet equipped with low precision pressure sensor matrix is already sufficient to detect subtle activities and identity of the person on the carpet. By a 0.4 m2 matrix of 32 × 32, 12 bit pressure sensors, we achieve 78.7% accuracy for 11 test subjects performing 7 subtle activities (open 7 different drawers or cabinet doors) and 88.6% accuracy in recognizing who has performed the activities. We thus see the potential of using a single carpet as a unified approach in houses to detect how inhabitants interact with the furniture without attaching different sensors onto each single furniture.

In this paper we present a compact and versatile acquisition platform designed by modularized method, which is useful and flexible for various sensing applications (such as mechanic detection and ubiquitous computing) thanks to the combination of high resolution analog-to-digital converter (ADC), low power consumption and exchangeable wireless transmission. The 31*45 mm² module supports simultaneous acquisition on maximum 8 channels with an effective number of bits (ENOB) of 20.9 at most and a sampling rate of up to 32KSPS each. Also we implement two popular modes of wireless transmission featuring high data rate and low power, namely a transmission rate of 21.3Kbps using Bluetooth low energy (BLE, power consumption 30mW) and 263Kbps using WiFi (723mW). Additionally, two typical wireless sensing applications, floor vibration detection and electrocardiograph (ECG) signal acquisition, are tested to demonstrate the performance of this platform. Results show that the characteristic of this platform can primely satisfy the requirements of these applications featuring high resolution, low power and high-efficient wireless data transmission, which is greatly helpful for exploring-stage experiments in wireless sensing application cases.

In this paper, we investigate how much information about user activity can be extracted from simple pressure sensors mounted under the legs of a chair. We show that it is possible to detect not only different postures (0.826 accuracy for 5 subjects and 7 We build on previous work [5] that demonstrated, in simple isolated experiments, how head and neck related events (eg swallowing, head motion) can be detected using an unobtrusive, textile capacitive sensor integrated in a collar like neckband. We have now developed a 2nd generation that allows long term recording in real life environments in conjunction with a low power Bluetooth enabled smart phone. It allows the system to move from the detection of individual swallows which is too unreliable for practical applications to an analysis of the statistical distribution of swallow frequency. Such an analysis allows the detection of" nutrition events" such as having lunch or breakfast. It also allows us to see the general level of activity and distinguish between just being absolutely quiet (no motion) and sleeping. The neckband can be useful in a variety of applications such as cognitive disease monitoring and elderly care. ) but also subtle hand and head related actions like typing and nodding (0.880 accuracy for 5 subjects and 5 classes). Combining features related to postures and such simple actions, we can detect high-level activities such as working on a PC, watching a movie or eating in a continuous, real-life data stream. In a dataset of 105.6 hours recorded from 4 subjects, we achieved 0.783 accuracy for 7 classes.