In recent years, the use of computer systems for identifying people has become increasingly popular. Clear facial images and detailed facial features are often provided to the system to improve identification accuracy, but if they are misused, privacy can be compromised. The authors are currently building a next-generation video blog system that can do the following: capture images of passersby with cameras, blur the facial images of all the people, remove the blurring for specific people who have registered their facial images for public viewing, and then upload them as video blogs in real time in a hands-free manner. By comparing the similarities between the obscured data in the user's stored data and the data captured by the camera and immediately obscured and sent to the edge server, the authors devised a system that enables person identification but does not pass clear facial images and features to the system. The authors focused on the possibility of using noise strength added to features and random seeding of features as a common quasi-encryption key to protect privacy. This paper assesses the extent to which the intensity of noise in this system obscures person identification. The results show that privacy can be protected by adding epsilon noise below a certain strength to the features, and the noise strength and the random number seed for noise generation can be used as a common quasi-cryptographic key.

In recent years, there has been an urgent need to improve safety and efficiency to avoid dangerous situations. In this paper, the authors proposes a visualization system of near-future information by augmented reality (AR) to avoid dangerous situations. The proposed system utilizes the position and speed information of the user and obstacles, predicts crashes, and recommends avoidance actions such as changing course or stopping. Experiments were conducted under different conditions by the implemented system to determine how many seconds into the future it would be safer for the user and easier to take avoidance action. The experiment results showed that the avoidance rate was almost the same for 2 seconds or longer, but the best rate was for 3 seconds ahead.

In recent years, the rapid growth of the logistics and delivery industry has increased the demand for safe and efficient delivery planning. However, conventional navigation services do not take road and driver characteristics into account and recommend the same delivery route to every delivery person under the same settings. In this paper, the authors propose a safe and efficient delivery plan based on road characteristics by considering the risk and distance of each road segment estimated using telematics data. First, the proposed method uses clustering to determine the destinations to be handled by delivery personnel, and then uses a genetic algorithm to determine the route and order of delivery. The proposed method was simulated on two different road networks to evaluate its usefulness. The simulation results show that the proposed method can develop a delivery plan that satisfies the specific conditions and needs of delivery personnel and considers both safety and efficiency.

Recent technological advances in Virtual Reality (VR) and Augmented Reality (AR) enable users to experience a high-quality virtual world. Using VR to experience the virtual world, the user's entire view becomes the virtual world, and the user's physical movement is generally limited because the user cannot see the surrounding situation in the real world. Using AR to experience the virtual world, we generally use special sensors such as LiDAR to detect the real space and superimpose the virtual world on the real space. However, it is difficult for devices without such special sensors to detect real space and superimpose a virtual world at an appropriate position. This study proposes two methods for replacing the background: a method using depth estimation and a method using semantic segmentation. This study also confirmed that the system can be used with sufficient removal accuracy and response time by using appropriate image size for the environment and that a safe and highly immersive virtual world experience can be achieved.

Route recommendation services have been widely used due to the spread of mobile devices such as smartphones. However, conventional route recommendation services often recommend difficult and unsafe routes which require skilled driving techniques because conventional services aim to recommend the shortest route considering the travel distance and time. In a previous study, we proposed the method of route recommendation that minimizes the accident rate by estimating the accident rate for each road segment according to the characteristics of the driver. The Dijkstra algorithm used for route selection could only consider one link weight, so we obtained some routes that were unnecessarily increasing the route length. Therefore, in this study, we applied a genetic algorithm to route selection and set a threshold for distance in the constraints. In the simulation, it was confirmed that it can recommend safe routes that reduce the accident rate satisfying the constraints by changing the thresholds by parameters and evaluating the results.

In this paper, the authors propose a real-time AR visualization system that enables the user to visually grasp the future snow-covered situation at the current location, aiming to support residents and visitors in heavy snow areas. The proposed system generates snow-covered spatial information that reflects the snow-covered situation on spatial information in the real world sensed by 3D LiDAR, and composes it on the video image captured by a camera. This paper describes a lightweight method of spatial information processing by reducing the amount of spatial data transmission.

In Japan, tsunamis and landslides are frequently caused as secondary disasters after earthquakes, and early evacuation is an important issue to reduce the damage. In this study, we propose a method to reduce the time required to complete evacuation using social graphs and to reduce the number of messages for evacuation based on geographical information. We propose a method to speed up the evacuation completion time and reduce the number of messages by setting an evacuation target area and evacuating only the targets (nodes) in the area. n Relaying an evacuation order only when there is a node in the evacuation area ahead of the node, the method speeds up the evacuation completion time and reduces the number of messages. The evaluation results confirmed that the proposed method was able to reduce the number of messages in the evacuation while maintaining a faster evacuation completion time compared to the case where all nodes outside the evacuation target area relayed the evacuation order and the case where no relay was performed.

Route recommendation services have been widely used due to the spread of mobile devices such as smartphones. However, conventional route recommendation services often rec-ommend difficult and unsafe routes which require skilled driving techniques because conventional services aim to recommend the shortest route considering the travel distance and time. Therefore, in this paper, we propose a safe route recommendation method. The proposed method estimates the accident rate for each road section based on driver characteristics from telematics data and recommends routes that minimize the estimated accident rate. A binary classification model is generated by machine learning from the acquired data, and the objective variable is the presence or absence of accidents. The features of the driver are input to the generated model and the accident rate is estimated for the case where that driver travels that road section. Simulation evaluation confirmed that the proposed method can recommend routes with lower accident rates than the distance-minimizing route recommendation method.

In this paper, we propose a tree construction method to collect and composite distributed videos on the live broad-casting system called “same world broadcasting system”. The same world broadcasting system generates and delivers a video in which video objects captured at the different places are displayed as if they were located in the same world. Our tree construction method employs an “occlusion graph” that represents the overlap relations of the video objects to give a heuristic solution to find a tree of network structure to collect and composite videos which provides a high-quality video in heterogeneous environments. We implemented the tree construction algorithm and conducted computer simulations considering different scenarios. Our simulation results show that our proposal can find a good collection tree within a practical amount of time.

Recent technological advances in Virtual Reality (VR) and Augmented Reality (AR) enable users to experience a high-quality virtual world. In VR applications, the user's physical movement is generally restricted because the situation around the real world cannot be seen. AR allows users to experience virtual worlds without restrictions on physical movement, but the extent to which they are replaced as virtual worlds is limited. In this research, assuming the use of smartphones and tablet devices, a partial virtual world system is implemented by removing only the background part from the real-time real-world image taken by the camera and replacing it with a virtual background.

Many systems and services for VR spaces use avatars to represent the appearances of each user. In those systems and services, the users communicate with each other via their avatars. Currently, various techniques and researches have been studied for avatar appearance. However, the existing techniques and studies mainly regard avatars as the self-expression of each user and do not mention the influences on communication from others. In communication content, avatars are one of self-expression of each user and one of the communication tools. Therefore, this paper describes an investigation of how avatar appearances affect communication from others. The result from the research showed that the degree of deformation of avatars affects communication from others in various situations such as informal and formal situations.

Due to the sensor devices price reduction and improvement of network connection quality, IoT(Internet of Things) is rapidly growing these days. IoT handles massive data that deeply related to time or location. Many peer-to-peer approaches that optimized this characteristic are existing. We have worked for a structured and virtualized overlay network that can efficiently handle time-consecutive specific interval queries. But this method had an unsolved problem that causes more load to the whole network than usual. In this work, we present an optimized routing method for an interval-queryable overlay network. To reduce redundant traffic, we expand the routing process to the physical node level. The simulation shows this method can reduce the load of the network or each node.

Recently, live video broadcasting is frequently used due to the COVID-19 pandemic. More realistic and comfortable experiments can be realized if multiple real-time videos from different places are displayed on a single screen. This paper calls such live broadcasting "same world broadcasting" as if all objects exist in the same space. In this paper, we propose a distributed system for the same world broadcasting. We evaluate the proposed system using the experimental implementation in a LAN environment and an open testbed environment. The experiment results show that the proposed system can reduce the processing time to collect and compose many real-time videos.

Various road pricing systems have been researched and implemented in many countries to reduce traffic congestion. The authors have also proposed a road pricing method based on demand distribution. The proposed method uses demand information to solve the combinatorial optimization problem for maximizing the total income of road providers under the constraint of specific traffic capacities. However, the previous work takes a long time to determine prices because it searches all combinations of departure and arrival places for appropriate prices. The searched combinations include unavailable routes. Therefore, this paper presents a complexity reduction method for the existing road pricing method based on demand distribution. The proposed method employs a technique of the compressed sparse row (CSR). The authors evaluated the proposed method and confirmed that the proposed method can reduce the calculation time in various road models.

This paper proposes "WoT-based Logical Sensor Architecture (WLSA)," a novel virtual sensor architecture on the basis of Web of Thing (WoT) that enables IoT applications to treat the sensor data and processing results in a uniform way in edge computing environments. WLSA reduces the computing and network resources required for the IoT applications by reusing the processing results in multiple data flows. This paper also proposes a data flow transforming (DFT) algorithm, in which a subset of the data flow is automatically replaced with a process that reuses the result of running processes so that the application developers do not need to conscious of sharing the processing result. We have implemented a prototype system and an object detection application based on WLSA using Node-RED data flow processing framework in an edge computing environment and evaluated its effectiveness. The evaluation results showed that WLSA reduced the execution time with less usage of computing resources. The prototype could accommodate 20 applications with less than 100 ms response time under 4% CPU usage.

Due to the recent popularization of live video distributions, live video distributors such as YouTubers have attracted great attention. Some of them shoot videos of themselves using smartphones while moving and distribute the videos via the Internet. One of the major consideration points of live video distributions on the Internet is trust between distributors and viewers, since various public concerns such as threats or attacks to distributors occur in live video distribution situations with no trust. To solve these concerns and enable safety live video distributions on the Internet, we propose a novel live video distribution architecture. In trusty situations, the systems under our proposed architecture add more video captions to give various information to the viewers. Otherwise, the systems add more effects to hide information. We also define three policies for distributors to manage trust.

In IoT (Internet of Things) systems, various devices such as cameras and sensors generate a huge stream data and are cyclically send them to processing computers via computer networks. The transaction rates for processing such stream data is one of the main performances for some IoT applications. Some schemes control transaction intervals to improve transaction rates. Most of them adopt a simple approach in that processing computers request the changes of transaction intervals to data sources every a fixed number of transactions. This approach unfortunately causes the drawback of unfair (inconstant) transaction intervals. Most of IoT applications prefer constant transaction intervals, e.g., a constant frame rate for video applications, since their constantness enable stable processing. In this paper, we propose a method to dynamically determine transaction intervals based on transaction rates. In our proposed method, the processing computer for an IoT application changes the transaction interval when the average value for some previous transactions largely changes. We measured the transaction rate of our proposed method in our developed simulator and confirmed that our proposed method can give a higher fairness of the transaction rates compared with existing methods by using appropriate parameters.

Internet live broadcasting services such as USTREAM, TwitCasting have become popular. These services give broadcasters an ability which can broadcast videos via the Internet. By adding video effects and audio effects, live streaming services can realize as if broadcasters were in a different world. In this paper, we call this kind of live streaming `Different Worlds Broadcasting'. However if many viewers request the effect in a short period of time or use some heavy load effect, it takes much time to render those effects and the frame rate gets worse. Therefore, currently users can use only small load effects such as superimposing small firework effect on the video, adding simple echo effect on the sound. Rendering effects in a short time, users can add more effects or the frame rate cat get better. In this paper, we propose a distributed video processing system for Internet live broadcasting services with video effect and audio effect. The proposed system take a video by broadcaster's PC or smartphone and transmit it to effect servers to add various effects. Effected video stream is transmitted to the broadcast server of the Internet live broadcasting service.

In this paper, we propose a lightweight multi-receiver encryption scheme for the device to device communications on Internet of Things (IoT) applications. In order for the individual user to control the disclosure range of his/her own data directly and to prevent sensitive personal data disclosure to the trusted third party, the proposed scheme uses device-generated public keys. For mutual authentication, third party generates Schnorr-like lightweight identity-based partial private keys for users. The proposed scheme provides source authentication, message integrity, replay-attack prevention and implicit user authentication. In addition to more security properties, computation expensive pairing operations are eliminated to achieve less time usage for both sender and receiver, which is favourable property for IoT applications. In this paper, we showed a proof of security of our scheme, computational cost comparison and experimental performance evaluations. We implemented our proposed scheme on real embedded Android devices and confirmed that it achieves less time cost for both encryption and decryption comparing with the existing most efficient certificate-based multi-receiver encryption scheme and certificateless multi-receiver encryption scheme.

The recent development of stream delivery technology made Internet broadcasting services popular. In live Internet broadcasting, broadcasters or viewers often add video or sound effects (we call this kind of live broadcasting, `different worlds broadcasting'). Therefore, we have proposed a distributed Internet live broadcasting system with video and audio effects. In this paper, we describe the implementation of our proposed system and its demonstration system using real-time facial animation.

Recently, sensor devices such as temperature sensors or cameras are used for various purposes. These devices sometimes construct ad hoc networks, and a grid topology is one of their typical network topologies. Describing processing by rules is suitable for them since the system can flexibly change their processing according to their purposes. By relieving their rule processing loads and communication delay, the system can be applied for many applications. However, there is no rule processing schemes to realize those. In this paper, we propose a rule processing scheme using the Rete algorithm. Our proposed scheme can realize rule processing reducing redundant loads and delay for many applications.

Due to the prevalence of sensors such as live cameras or environmental sensors, sensor data stream delivery, which requires continuous and cyclic data delivery attracts great attention. For sensor data stream delivery, various communication loads balancing techniques have been studied since the load of the sensor data source become high to accommodate large number of clients. However, these studies assume only the requests that have the same collection cycle, which is not enough for the actual applications. In this paper, we propose a sensor data stream delivery system with communication loads balancing for heterogeneous collection cycle requests. The proposed system distributes the loads by re-delivering the sensor data that are requested by other clients with different collection cycles but have common cycles.