Recent years have seen remarkable progress in human pose estimation. However, manual annotation of keypoints remains tedious and imprecise. To alleviate this problem, this paper proposes a novel method called Multi-Scale Contrastive Learning (MSCL). This method uses a siamese network structure with upper and lower branches that capture diffirent views of the same image. Each branch uses a backbone network to extract image representations, employing multi-scale feature vectors to capture information. These feature vectors are then passed through an enhanced feature pyramid for fusion, producing more robust feature representations. The feature vectors are then further encoded by mapping and prediction heads to predict the feature vector of another view. Using negative cosine similarity between vectors as a loss function, the backbone network is pre-trained on a large-scale unlabeled dataset, enhancing its capacity to extract visual representations. Finally, transfer learning is performed on a small amount of labelled data for the pose estimation task. Experiments on COCO datasets show significant improvements in Average Precision (AP) of 1.8%, 0.9%, and 1.2% with 1%, 5%, and 10% labelled data on COCO. In addition, the Percentage of Correct Keypoints (PCK) improves by 0.5% on MPII&AIC, outperforming mainstream contrastive learning methods.

In recent years, location-based technologies for ubiquitous environments have aimed to realize services tailored to each purpose based on information about an individual's current location. To establish such advanced location-based services, an estimation technology that can accurately recognize and predict the movements of people and objects is necessary. Although global positioning system (GPS) has already been used as a standard for outdoor positioning technology and many services have been realized, several techniques using conventional wireless sensors such as Wi-Fi, RFID, and Bluetooth have been considered for indoor positioning technology. However, conventional wireless indoor positioning is prone to the effects of noise, and the large range of estimated indoor locations makes it difficult to identify human activities precisely. We propose a method to mine user activity patterns from time-series data of user's locationss in an indoor environment using ultra-wideband (UWB) sensors. An UWB sensor is useful for indoor positioning due to its high noise immunity and measurement accuracy, however, to our knowledge, estimation and prediction of human indoor activities using UWB sensors have not yet been addressed. The proposed method consists of three steps: 1) obtaining time-series data of the user's location using a UWB sensor attached to the user, and then estimating the areas where the user has stayed; 2) associating each area of the user's stay with a nearby landmark of activity and assigning indoor activities; and 3) mining the user's activity patterns based on the user's indoor activities and their transitions. We conducted experiments to evaluate the proposed method by investigating the accuracy of estimating the user's area of stay using a UWB sensor and observing the results of activity pattern mining applied to actual laboratory members over 30-days. The results showed that the proposed method is superior to a comparison method, Time-based clustering algorithm, in estimating the stay areas precisely, and that it is possible to reveal the user's activity patterns appropriately in the actual environment.

In recent years, cyber attacks against infrastructure have become more serious. Unfortunately, infrastructures with vulnerable remote management devices, which allow attackers to control the infrastructure, have been reported. Targeted attacks against infrastructure are conducted manually by human attackers rather than automated scripts. Here, open questions are how often the attacks against such infrastructure happen and what attackers do after intrusions. In this empirical study, we observe the accesses, including attacks and security investigation activities, using the customized infrastructure honeypot. The proposed honeypot comprises (1) a platform that easily deploys real devices as honeypots, (2) a mechanism to increase the number of fictional facilities by changing the displayed facility names on the WebUI for each honeypot instance, (3) an interaction mechanism with visitors to infer their purpose, and (4) tracking mechanisms to identify visitors for long-term activities. We implemented and deployed the honeypot for 31 months. Our honeypot observed critical operations, such as changing configurations of a remote management device. We also observed long-term access to WebUI and Telnet service of the honeypot.

In recent years, microwave wireless power transfer (WPT) has attracted considerable attention due to the increasing demand for various sensors and Internet of Things (IoT) applications. Microwave WPT requires technology that can detect and avoid human bodies in the transmission path. Using a phantom is essential for developing such technology in terms of standardization and human body protection from electromagnetic radiation. In this study, a simple and lightweight phantom was developed focusing on its radar cross-section (RCS) to evaluate human body avoidance technology for use in microwave WPT systems. The developed phantom's RCS is comparable to that of the human body.

An enzymatic biofuel cell (BFC) that uses lactic acid in human sweat as fuel to generate electricity is an attractive power source for wearable devices. A BFC capable of generating electricity with human sweat has been developed. It comprised a flexible tattoo seal type battery with silver oxide vapor deposited on a flexible material and conductive carbon nanotubes printed on it. The anode and cathode in this battery were arranged in a plane (planar type). This work proposes a thin laminated enzymatic BFC by inserting a cellulose nanofiber (CNF) sheet between two electrodes to absorb human sweat (stack-type). Optimization of the anode and changing the arrangement of electrodes from planar to stack type improved the output and battery life. The stack type is 43.20μW / cm2 at 180mV, which is 1.25 times the maximum power density of the planar type.

We propose a framework for the integration of heterogeneous networks in human pose estimation (HPE) with the aim of balancing accuracy and computational complexity. Although many existing methods can improve the accuracy of HPE using multiple frames in videos, they also increase the computational complexity. The key difference here is that the proposed heterogeneous framework has various networks for different types of frames, while existing methods use the same networks for all frames. In particular, we propose to divide the video frames into two types, including key frames and non-key frames, and adopt three networks including slow networks, fast networks, and transfer networks in our heterogeneous framework. For key frames, a slow network is used that has high accuracy but high computational complexity. For non-key frames that follow a key frame, we propose to warp the heatmap of a slow network from a key frame via a transfer network and fuse it with a fast network that has low accuracy but low computational complexity. Furthermore, when extending to the usage of long-term frames where a large number of non-key frames follow a key frame, the temporal correlation decreases. Therefore, when necessary, we use an additional transfer network that warps the heatmap from a neighboring non-key frame. The experimental results on PoseTrack 2017 and PoseTrack 2018 datasets demonstrate that the proposed FSPose achieves a better balance between accuracy and computational complexity than the competitor method. Our source code is available at https://github.com/Fenax79/fspose.

Contemporary development projects benefit from code review as it improves the quality of a project. Large ecosystems of inter-dependent projects like OpenStack generate a large number of reviews, which poses new challenges for collaboration (improving patches, fixing defects). Review tools allow developers to link between patches, to indicate patch dependency, competing solutions, or provide broader context. We hypothesize that such patch linkage may also simulate cross-collaboration. With a case study of OpenStack, we take a first step to explore collaborations that occur after a patch linkage was posted between two patches (i.e., cross-patch collaboration). Our empirical results show that although patch linkage that requests collaboration is relatively less prevalent, the probability of collaboration is relatively higher. Interestingly, the results also show that collaborative contributions via patch linkage are non-trivial, i.e, contributions can affect the review outcome (such as voting) or even improve the patch (i.e., revising). This work opens up future directions to understand barriers and opportunities related to this new kind of collaboration, that assists with code review and development tasks in large ecosystems.

Machine-to-Machine (M2M) communication plays a pivotal role in the evolution of Internet of Things (IoT). Cellular networks are considered to be a key enabler for M2M communications, which are originally designed mainly for Human-to-Human (H2H) communications. The introduction of M2M users will cause a series of problems to traditional H2H users, i.e., interference between various traffic. Resource allocation is an effective solution to these problems. In this paper, we consider a shared resource block (RB) and power allocation in an H2H/M2M coexistence scenario, where M2M users are subdivided into delay-tolerant and delay-sensitive types. We first model the RB-power allocation problem as maximization of capacity under Quality-of-Service (QoS) constraints of different types of traffic. Then, a learning framework is introduced, wherein a complex agent is built from simpler subagents, which provides the basis for distributed deployment scheme. Further, we proposed distributed Q-learning based autonomous RB-power allocation algorithm (DQ-ARPA), which enables the machine type network gateways (MTCG) as agents to learn the wireless environment and choose the RB-power autonomously to maximize M2M pairs' capacity while ensuring the QoS requirements of critical services. Simulation results indicates that with an appropriate reward design, our proposed scheme succeeds in reducing the impact of delay-tolerant machine type users on critical services in terms of SINR thresholds and outage ratios.

In recent years, High-Altitude Platform Station (HAPS) has become the most interesting topic for next generation mobile communication systems, because platforms such as Unmanned Aerial Vehicles (UAVs), balloons, airships can provide ultra-wide coverage, up to 200km in diameter, from altitudes of around 20 km. It also offers resiliency to damage caused by disasters and so ensures the stability and reliability of mobile communications. In order to further integrate HAPS with existing terrestrial mobile communication networks in providing mobile services to users, radio wave propagation models such as terrain, vegetation loss, human shielding loss, building entry loss, urban/suburban areas must be taken into consideration when designing HAPS-based cell configurations. This paper proposes a human body shielding propagation loss model that considers the basic signal attenuation by the human body at high elevation angles. It also analyzes the effect of changes in actual urban/suburban environments due to the arrival of multipath radio waves for HAPS communications in the frequency range of 0.7 to 3.3GHz. Measurements in actual urban/rural environments in Japan and actual stratospheric base station measurements in Kenya are carried out to confirm the validity of the proposed model. Since the measured results agree well with the results predicted by the proposed model, the model is good enough to provide estimates of human loss in various environments.

The road space rationing (RSR) method regulates a period in which a user group can make telephone calls in order to decrease the call attempt rate and induce calling parties to shorten their calls during disaster congestion. This paper investigates what settings of this indirect control induce more self-restraint and how the settings change calling parties' behavior using experimental psychology. Our experiments revealed that the length of the regulated period differently affected calling parties' behavior (call duration and call attempt rate) and indicated that the 60-min RSR method (i.e., 10 six-min periods) is the most effective setting against disaster congestion.

A method to generate color palettes from images is proposed. Here, deep neural networks (DNN) are utilized in order to consider human perception. Two aspects of human perception are considered; one is attention to image, and the other is human preference for colors. This method first extracts N regions with dominant color categories from the image considering human attention. Here, N is the number of colors in a color palette. Then, the representative color is obtained from each region considering the human preference for color. Two deep neural-net systems are adopted here, one is for estimating the image area which attracts human attention, and the other is for estimating human preferable colors from image regions to obtain representative colors. The former is trained with target images obtained by an eye tracker, and the latter is trained with dataset of color selection by human. Objective and subjective evaluation is performed to show high performance of the proposed system compared with conventional methods.

The effect of provision of “Neither-Good-Nor-Bad” (NGNB) information on the perceived trustworthiness of agents has been investigated in previous studies. The experimental results have revealed several conditions under which the provision of NGNB information works effectively to make users perceive greater trust of agents. However, the experiments in question were carried out in a situation in which a user is able to choose, with the agent's advice, one of a limited number of options. In practical problems, we are often at a loss as to which to choose because there are too many possible options and it is not easy to narrow them down. Furthermore, in the above-mentioned previous studies, it was easy to predict the size of profits that a user would obtain because its pattern was also limited. This prompted us, in this paper, to investigate the effect of provision of NGNB information on the users' trust of agents under conditions where it appears to the users that numerous options are available. Our experimental results reveal that an agent that reliably provides NGNB information tends to gain greater user trust in a situation where it appears to the users that there are numerous options and their consequences, and it is not easy to predict the size of profits. However, in contradiction to the previous study, the results in this paper also reveal that stable provision of NGNB information in the context of numerous options is less effective in a situation where it is harder to obtain larger profits.

This paper proposes a method for cooperative multi-static Multiple Input Multiple Output (MIMO) radar that can estimate the number of targets. The purpose of this system is to monitor humans in an indoor environment. First, target positions within the estimation range are roughly detected by the Capon method and the mode vector corresponding to the detected positions is calculated. The mode vector is multiplied by the eigenvector to eliminate the virtual image. The spectrum of the evaluation function is calculated from the remaining positions, and the number of peaks in the spectrum is defined as the number of targets. Experiments carried out in an indoor environment confirm that the proposed method can estimate the number of targets with high accuracy.

The need for real-time use of human dynamics data is increasing. The technical requirements for this include improved databases for handling a large amount of data as well as highly accurate sensing of people's movements. A bitmap index format has been proposed for high-speed processing of data that spreads in a two-dimensional space. Using the same format is expected to provide a service that searches queries, reads out desired data, visualizes it, and analyzes it. In this study, we propose a coding format that enables human dynamics data to compress it in the target data size, in order to save data storage for successive increase of real-time human dynamics data. In the proposed method, the spatial population distribution, which is expressed by a probability distribution, is approximated and compressed using the one-pixel one-byte data format normally used for image coding. We utilize two kinds of approximation, which are accuracy of probability and precision of spatial location, in order to control the data size and the amount of information. For accuracy of probability, we propose a non-linear mapping method for the spatial distribution, and for precision of spatial location, we propose spatial scalable layered coding to refine the mesh level of the spatial distribution. Also, in order to enable additional detailed analysis, we propose another scalable layered coding that improves the accuracy of the distribution. We demonstrate through experiments that the proposed data approximation and coding format achieve sufficient approximation of spatial population distribution in the given condition of target data size.

We developed an 8.4-inch electrostatic-tactile touch display using a segmented-electrode array (30×20) as both tactile pixels and touch sensors. Each pixel can be excited independently so that the electrostatic-tactile touch display allows presenting real localized tactile textures in any shape. A driving scheme in which the tactile strength is independent of the grounding state of the human body by employing two-phased actuation was also proposed and demonstrated. Furthermore, tactile crosstalk was investigated to find it was due to the voltage fluctuation in the human body and it was diminished by applying the aforementioned driving scheme.

As characterizing videos simultaneously from spatial and temporal cues have been shown crucial for video processing, with the shortage of temporal information of soft assignment, the vector of locally aggregated descriptor (VLAD) should be considered as a suboptimal framework for learning the spatio-temporal video representation. With the development of attention mechanisms in natural language processing, in this work, we present a novel model with VLAD following spatio-temporal self-attention operations, named spatio-temporal self-attention weighted VLAD (ST-SAWVLAD). In particular, sequential convolutional feature maps extracted from two modalities i.e., RGB and Flow are receptively fed into the self-attention module to learn soft spatio-temporal assignments parameters, which enabling aggregate not only detailed spatial information but also fine motion information from successive video frames. In experiments, we evaluate ST-SAWVLAD by using competitive action recognition datasets, UCF101 and HMDB51, the results shcoutstanding performance. The source code is available at:https://github.com/badstones/st-sawvlad.

This paper introduces a novel skin-stretcher device for gently urging head rotation. The device pulls and/or pushes the skin on the user's neck by using servo motors. The user is induced to rotate his/her head based on the sensation caused by the local stretching of skin. This mechanism informs the user when and how much the head rotation is requested; however it does not force head rotation, i.e., it allows the user to ignore the stimuli and to maintain voluntary movements. We implemented a prototype device and analyzed the performance of the skin stretcher as a human-in-the-loop system. Experimental results define its fundamental characteristics, such as input-output gain, settling time, and other dynamic behaviors. Features are analyzed, for example, input-output gain is stable within the same installation condition, but various between users.

When using a gesture-based interface for pointing to targets on a wide screen, displaying a large pointer instead of a typical spot pattern reduces disturbance caused by measurement errors of user's pointing posture. However, it remains unclear why a large pointer helps facilitate easy pointing. To examine this issue, in this study we propose a mathematical model that formulates human pointing motions affected by a large pointer. Our idea is to describe the effect of the large pointer as human visual perception, because the user will perceive the pointer-target distance as being shorter than it actually is. We embedded this scheme, referred to as non-linear distance filter (NDF), into a typical feedback loop model designed to formulate human pointing motions. We also proposed a method to estimate NDF mapping from pointing trajectories, and used it to investigate the applicability of the model under three typical disturbance patterns: small vibration, smooth shift, and step signal. Experimental results demonstrated that the proposed NDF-based model could accurately reproduced actual pointing trajectories, achieving high similarity values of 0.89, 0.97, and 0.91 for the three respective disturbance patterns. The results indicate the applicability of the proposed method. In addition, we confirmed that the obtained NDF mappings suggested rationales for why a large pointer helps facilitate easy pointing.

This paper describes the effectiveness of the geometric multi-grid method in a current density analysis using a numerical human body model. The scalar potential finite difference (SPFD) method is used as a numerical method for analyzing the current density inside a human body due to contact with charged objects in a low-frequency band, and research related to methods to solve faster large-scale simultaneous equations based on the SPFD method has been conducted. In previous research, the block incomplete Cholesky conjugate gradients (ICCG) method is proposed as an effective method to solve the simultaneous equations faster. However, even though the block ICCG method is used, many iterations are still needed. Therefore, in this study, we focus on the geometric multi-grid method as a method to solve the problem. We develop the geometric-multi-grid method and evaluate performances by comparing it with the block ICCG method in terms of computation time and the number of iterations. The results show that the number of iterations needed for the geometric multi-grid method is much less than that for the block ICCG method. In addition, the computation time is much shorter, depending on the number of threads and the number of coarse grids. Also, by using multi-color ordering, the parallel performance of the geometric multi-grid method can be greatly improved.

Reinforcement learning methods achieve performance superior to humans in a wide range of complex tasks and uncertain environments. However, high performance is not the sole metric for practical use such as in a game AI or autonomous driving. A highly efficient agent performs greedily and selfishly, and is thus inconvenient for surrounding users, hence a demand for human-like agents. Imitation learning reproduces the behavior of a human expert and builds a human-like agent. However, its performance is limited to the expert's. In this study, we propose a training scheme to construct a human-like and efficient agent via mixing reinforcement and imitation learning for discrete and continuous action space problems. The proposed hybrid agent achieves a higher performance than a strict imitation learning agent and exhibits more human-like behavior, which is measured via a human sensitivity test.