The graph structures generated by natural or simple heuristics often fail to represent the spatial correlations influenced by complex factors. Therefore, introducing graph structure learning (GSL) can enhance the graph neural network-based spatial interpolation models. However, the input features of the GSL module are systematically unbalanced in spatial interpolation tasks. For example, many natural variables follow Gaussian- or gamma-distribution, and sensor spatial distributions are generally uneven. Thus, the GSL module must systematically incorporate corresponding solutions to avoid negatively impacting its generalization ability and degrading model performance. Our proposed model utilizes two encoders to embed feature deviations of node readings and centroid distance from preset distributions, respectively. Notably, these encoders are jointly optimized with other model components, and their generalization ability is improved through an adaptively adjustable information bottleneck. Consequently, the GSL module can offer a more robust graph structure by explicitly perceiving feature deviations in the input. Experimental results demonstrate that our model outperforms existing state-of-the-art baselines across multiple real-world datasets with diverse characteristics.

Recent trends in Wearable Human Activity Recognition (WHAR) have led to an unprecedented 42.9% increase in scholarly articles in 2022, underscoring the urgency for a comprehensive review to systematically categorize their varied research directions. Moreover, our analysis reveals that the contributions of current articles often deviate from the traditional stages of the human activity recognition pipeline, as established in prior literature. This misalignment suggests the necessity for an updated pipeline that more accurately reflects the intricacies and nuances of WHAR studies. In response, we review WHAR articles from 2021 to 2023 and introduce an innovative WHAR pipeline, emphasizing a research-focused approach. This new pipeline offers distinct advantages: it provides researchers with a clear and systematic categorization of WHAR articles, thereby enhancing understanding of the field. For practitioners, it facilitates the selection of customized methods for each stage, thereby optimizing final assembled model efficacy.

To this day, various approaches for providing local explanation of black box machine learning models have been introduced. Despite these efforts, existing methods suffer from deficiencies such as being difficult to comprehend, only considering one feature at a time and disregarding inter-feature dependencies, lacking meaningful values for each feature, or only highlighting features that support the model's decision. To overcome these drawbacks, this study presents a new approach to explain the predictions of any black box classifier, called Monte Carlo tree search for eXplainable Artificial Intelligence (McXai). It employs a reinforcement learning strategy and models the explanation generation as two distinct games. In the first game, the objective is to identify feature sets that support the model's decision, while in the second game, the aim is to find feature sets that lead to alternative decisions. The output is a human-friendly representation in the form of a tree structure, where each node represents a set of features to be examined, with less specific interpretations at the top of the tree. Our experiments demonstrate that the features identified by McXai are more insightful with regard to the classifications compared to traditional algorithm like LIME and Gram-cam. Furthermore, the ability to identify misleading features provides guidance towards improved robustness of the black box classifier.

User experience (UX) summarizes user perceptions and responses resulting from the interaction with a product, system, or service. The User Experience Questionnaire (UEQ) is one standardized instrument for measuring UX. With six scales, it identifies areas in which product improvements will have the highest impact. In this paper, we evaluate the reliability and validity of this questionnaire. The data of N = 1, 121 participants who interacted with one of 23 products indicated an acceptable to good reliability of all scales. The results show, however, that the scales were not independent of each other. Combining perspicuity, efficiency, and dependability to pragmatic aspects as well as novelty and stimulation to hedonic aspects of UX improved the model fit significantly. The systematic variations of product properties and correlations with the System Usability Scale (SUS) in a second experiment with N=499 participants supported the validity of these two factors. Practical implications of the results are discussed.

Reading and understanding source code is a major task in software development. Code comprehension depends on the quality of code, which is impacted by code structure and identifier naming. In this paper we empirically investigated whether longer but more descriptive identifier names improve code comprehension compared to short names, as they represent useful information in more detail. In a web-based study 88 Java developers were asked to locate a semantic defect in source code snippets. With descriptive identifier names, developers spent more time in the lines of code before the actual defect occurred and changed their reading direction less often, finding the semantic defect about 14% faster than with shorter but less descriptive identifier names. These effects disappeared when developers searched for a syntax error, i.e., when no in-depth understanding of the code was required. Interestingly, the style of identifier names had a clear impact on program comprehension for more experienced developers but not for less experienced developers.

Acoustic Doppler shift estimation is a cost-effective way to implement Human-Computer Interaction applications across existing smart devices such as smart phones and smart spekaers. However, due to the inherent uncertainty principle in the traditional time-frequency analysis, it remains challenging to profile motions accurately and timely. In this paper, phase offset in acoustic OFDM signal is leveraged for developing Phascope, a fine-grained, fast and flexible motion profiling scheme. We evaluate Phascope using simulation and experiment on COTS devices. Sub-millisecond response time is achieved for Phascope in our experiment. Besides, with optimal subcarrier selection and SNR of 30 dB over all subcarriers, Phascope can estimate motion speed of 0.1 m/s with 6.75% root-mean-square error (RMSE) compared to optimized FFT method.

This work presents the bPart, a highly integrated autonomous sensor platform for use with mobile phones and devices. It consists of a Bluetooth Low Energy (BLE) radio and several MEMS sensors, all integrated in a volume of less than 1cm³, including the battery. Aside from the wireless transceiver, the bPart features sensors for ambient illumination, 3D-acceleration, temperature and relative humidity. In addition, there is a button and a magnetic switch for binary input and a RGB-LED for user feedback. A secondary LED in the infrared spectrum enables camera-assisted identification and tracking of the node. Runtimes of several years are possible on the included CR2023 lithium coin cell, through the low energy radio, onboard power-conversion and low-power sleep modes. The latter is rated below 2µW and a single data packet consumes about 75µWs. Its low energy consumption makes the bPart suitable for operation with energy harvesting, which we have validated with a 33cm² solar cell in indoor lighting conditions.

Displaying three-dimensional content on a flat display is bound to reduce the impression of depth, particularly for mobile video see-trough augmented reality. Several applications in this domain can benefit from accurate depth perception, especially if there are contradictory depth cues, like occlusion in a x-ray visualization. The use of stereoscopy for this effect is already prevalent in head-mounted displays, but there is little research on the applicability for hand-held augmented reality. We have implemented such a prototype using an off-the-shelf smartphone equipped with a stereo camera and an autostereoscopic display. We designed and conducted an extensive user study to explore the effects of stereoscopic hand-held augmented reality on depth perception. The results show that in this scenario depth judgment is mostly influenced by monoscopic depth cues, but our system can improve positioning accuracy in challenging scenes.

This paper proposes using a 2D waveguide for a novel means of authentication in public Wi-Fi infrastructures. The design of the system is presented, and its practicability and usability is comparatively discussed with that of five other tag and context based authentication schemes, two of which have not been previously realized. In accordance with the presented application scenarios, all of the schemes were implemented in a platform-independent fashion built on web technology.

This paper presents a mobile, low-cost particulate matter sensing approach for the use in Participatory Sensing scenarios. It shows that cheap commercial off-the-shelf (COTS) dust sensors can be used in distributed or mobile personal measurement devices at a cost one to two orders of magnitude lower than that of current hand-held solutions, while reaching meaningful accuracy. We conducted a series of experiments to juxtapose the performance of a gauged high-accuracy measurement device and a cheap COTS sensor that we fitted on a Bluetooth-enabled sensor module that can be interconnected with a mobile phone. Calibration and processing procedures using multi-sensor data fusion are presented, that perform very well in lab situations and show practically relevant results in a realistic setting. An on-the-fly calibration correction step is proposed to address remaining issues by taking advantage of co-located measurements in Participatory Sensing scenarios. By sharing few measurement across devices, a high measurement accuracy can be achieved in mobile urban sensing applications, where devices join in an ad-hoc fashion. A performance evaluation was conducted by co-locating measurement devices with a municipal measurement station that monitors particulate matter in a European city, and simulations to evaluate the on-the-fly cross-device data processing have been done.

This paper introduces a fuzzy inference system, based on the Takagi-Sugeno-Kang model, to achieve efficient and reliable classification in the domain of ubiquitous computing, and in particular for smart or context-aware, sensor-augmented devices. As these are typically deployed in unpredictable environments and have a large amount of correlated sensor data, we propose to use a Gath-Geva clustering specification as well as a genetic algorithm approach to improve the model's robustness. Experiments on data from such a sensor-augmented device show that accuracy is boosted from 83% to 97% with these optimizations under normal conditions, and for more. challenging data from 54% to 79%.

This paper presents an experience report illustrating the design of the uPart tiny low-power sensor network platform: from the analysis phase over the definition of the application, design and construction of hardware, the implementation of the software and network to the application set-up. uPart sensor nodes were given away in the conference badge to 500 voluntary attendees of the Ubicomp 2005. In our demo application, uParts were able to recognize activities of attendees of the Ubicomp 2005 conference. Design was carried out under serve time and budget restrictions. The paper focuses on reporting design decisions and presents tech-nical details of uPart hardeware, firmware and applications. It also shows first qualitative experiences with the run of the system at the conference. The outcome of the paper is a general meta-guideline for designing sensor network systems under similar conditions.

Current sensor network platforms perform multiple processes including sensor sampling, communication, and various computational tasks. When deployed in unpredictable environments, complex schedules of those processes may arise. Typical sensor network qualities like periodic sampling of sensors, avoidance of process starvation and automatic energy management are required to be maintained in such situations. We propose a scheduling framework for senor nodes consisting of a scheduler, a dispatcher and a controller for an on-line adaptation of the process execution. The key components are a controller and an enhanced dispatcher implementing strategies like jitter correction and starvation avoidance. Further, the framework is aware of the energy consumption of sensors. We show that our controlled scheduling framework performs significantly better than a non-controlled single scheduler in unpredictable environments. Our proposed measures are efficient to implement. Results are shown by extensive simulations and a first implementation on the Particle Computer platform.