The Urban Heat Island (UHI) phenomenon is intensifying in urban areas, yet existing high-resolution temperature estimation methods, including LULC-based machine learning and kriging, do not adequately account for solar radiation and wind effects on microclimate, limiting daytime accuracy. We generate meter-scale radiation and wind maps from regional meteorological data using physical simulation software, and use them as features to estimate block-level temperature distributions. Field experiments on a university campus demonstrate that the proposed method improves estimation accuracy compared to conventional approaches, particularly during daytime.

Software-defined networking (SDN) enables flexible, programmable networks, but in large deployments, poor controller placement can raise latency and energy use. This paper presents a practical in-band SDN model and an optimization method that jointly accounts for control-plane delay and device power consumption to place controllers and configure switches. We formulate the problem as a binary integer programming (BIP) that decides controller locations, which switches remain active, and port bit-rates and routes. Experiments on a large WAN topology (Janos-US) show the approach can cut total network energy by up to 15% while keeping control-plane delays within required bounds, offering network operators a straightforward way to trade responsiveness for energy savings.

In Multi-Object Tracking (MOT) based on trackingby-detection, appearance features as well as position features of objects detected in each frame are used to associate objects across frames, but using appearance features is computationally expensive. We propose a method to speed up MOT by selectively using appearance features only in necessary situations. This method evaluates the states of objects, and applies a supervised decision tree to efficiently select the situations where appearance features need to be used. Experiments on MOT20 dataset confirmed that our method can speed up MOT by 1.2 times while achieving tracking accuracy equal to or better than the existing method.

Regardless of the widespread use of digital games, non-digital games are still popular, because of their tactile and spatial interaction in real spaces. Recently, Augmented Reality (AR) has also been incorporated into these games. This study focuses on card games, and proposes an AR Card game that preserves the original interaction and provides digital information, consequently enhancing their entertainment value. This paper outlines the development of a game incorporating AR-based functionality into the card game “Old Maid.”

Hand-object interactions play a central role in human activity. Analyzing them in depth requires not only identifying such actions in a scene, but also detecting when a hand is touching and affecting an object. We propose a novel method to detect hands moving objects, a reliable clue that a hand is touching and affecting an object. Our method focuses on the fact that an object being moved by a hand exhibits movements similar to those of the forearm, and detects the hand moving an object by determining whether movements similar to the forearm occur in the hand surrounding region. In this method, the person’s skeleton and motion obtained from the video are integrated into a feature vector that is then used to determine the movements in the hand surrounding region, allowing efficient utilization of multimodal information for the detection.

Non-digital games provide the tactile and spatial interaction of moving objects in real space, an aspect absent in digital games. This study considers card games as interactive media and introduces an “AR card game” using an optical see-through HMD. The proposed “AR Card Game” leverages AR to enhance digital information within the actual card game, preserving its inherent interactivity, and augmenting the entertainment value by controlling player information. This paper outlines the development of a card game incorporating AR-based functionality into the game "Old Maid".

Rapid urbanization and climate change have increased the frequency of extreme weather events. Understanding fine-grained weather conditions within cities is crucial for effective urban design and behavior change. However, extensive sensor networks are needed to achieve this granularity. This study proposes a method to estimate weather conditions in unsensed areas using Kriging, a probabilistic approach that combines data from fixed and mobile sensors. Experimental results demonstrate that the proposed approach can increase the estimation accuracy by approximately 0.93 °C compared to the existing inverse distance weighting method.

A lot of the devices in the Internet of Things are sensors responsible for capturing environmental information and relaying it to a system or network. Oftentimes, it is important to know the location of these sensors to better contextualize the information received from them. However, the sensors are purposely simple and cheap, meaning that conventional localization techniques, such as global navigation satellite systems are not feasible. Research has been done on using unmanned aerial vehicles to estimate the location of the sensors, but issues with signal strength fluctuation and location approximation because of it are still prevalent. In this work, we propose a new method for estimating the location of sensors by exploiting the characteristics of radio wave signal propagation and creating a new flight path design, a solution to minimize the impact of variation in measurements, a novel candidate point generation through signal strength analysis, and a method to find the location based on the candidates. Through a thought-out experiment, we show that the proposed algorithm is overall significantly better than the existing solution when it comes to identifying the location of outdoor sensors.

The action of a person moving an object with their hands plays a central role in human-object interaction. In this article, we propose a novel method to detect the action of moving an object with the hands in RGB-D video. Our method uses 2D skeleton information and 3D motion (scene) flow obtained from RGB-D video to analyze the movements around the hand, and determines that the hand moves an object when movements similar to those of the forearm occur around the hand. By effectively utilizing different information available from RGB-D video, this method is expected to be able to efficiently detect actions of moving objects with hands in various directions.

The action of moving an object with a hand is one of the main activities in human-object interaction. This paper proposes a novel method to detect human hands moving objects in video. The proposed method analyzes the motion information in the area around the extracted forearm tip and determines that the hand is moving an object if there is a part showing the similar motion as something moved by hand in the surrounding area. Since this method detects hands moving objects based on features that can be acquired without extracting the object regions from video, it is expected to be applied to various situations.

The market for Domotics, most commonly known as Home Automation, has exponentially grown in the past few years. In brief, Domotics allows users to monitor and control various devices/appliances within a smart home. From entertainment to security, several devices have become more accessible. However, the underlying network infrastructure in this environment has remained the same despite the recent revolution brought about by Software-Defined Networking (SDN).For the past decade, SDN has shown great potential for providing more network flexibility and programmability in both the control and data plane. However, most of the use cases are thought for enterprise and Data-Center Networks, while its application closer to the end-users (i.e., Home environments) is still under-exploited. Therefore, this research introduces the Network Operations for Domotics (NeO-Domotics) notion, which uses Next-Generation SDN components (e.g., P4, NG-SDN interfaces) in the context of Home Automation. This paper describes the overall design and showcases preliminary use cases using a Raspberry Pi.

It has been difficult to effectively develop performance skills in guitar playing until now, and attempts have been made to utilize information technology to support skill development. In recent years, systems have been developed that apply Extended Reality(XR) technology to superimpose learning support information on the real instrument periphery to achieve more advanced support. However, existing systems for guitar performance training focus only on whether the strings are pressed at the correct position on the fingerboard, and it is difficult to provide comprehensive support for a variety of playing skills. In this study, we propose a performance skill training support system based on XR technology to support more advanced guitar playing techniques. In addition to the presentation of string pressing positions in the conventional system, the proposed system aims to support the acquisition of more practical playing skills by implementing a function to evaluate playing skills in real-time using acoustic analysis. This paper presents an overview of the study, the initial implementation of the proposed system, and the results of user evaluations.

Software-Defined Networking (SDN) separates the control from the data plane in the network infrastructure to flexibly and efficiently manage the resources. Since the controller is the core of the control plane, multiple instances need to be placed within the network to increase performance, reliability, and fault tolerance, among others. However, deciding the number and where to place the allotted controllers is not trivial; it is known as the Controller Placement Problem (CPP). This paper considers the energy-efficiency aspect of the network infrastructure to solve the CPP. To do so, we designed a model that includes delay constraints, whose effectivity was assessed and confirmed to save about 20-30% more energy than a model that only considers the number of controllers. Moreover, it is confirmed that execution time can be decreased by 100 times while keeping the energy consumption low by reducing the number of pre-calculated paths between the SDN switches and the controllers.

We propose card games in which various information is projected onto real cards by using Augmented Reality (AR) technology with an optical see-through head mounted display (HMD). In this paper, based on the results of preliminary experiments using a prototype of the initial implementation of this system, we discuss the requirements of the AR cards used in this system and the functional design of the AR card games based on these requirements.

AC losses in a high temperature superconducting (HTS) coil are experimentally evaluated. Double-pancake (DP) coils with turn-to-turn electrical insulations are wound using bundle conductors formed with two pieces of rare-earth-based coated conductors located face-to-face without insulation to improve thermal stability. Five DP coils are stacked and subsequently four copper plates are soldered between the adjacent DP coils to form a small-size HTS coil for AC loss measurements. The fabricated HTS coil is immersed in liquid nitrogen. Before measuring the AC losses, the contact resistances between the DP coils are observed in DC operations of the HTS coil at first. After that, the total losses including the Joule losses in the joints between the DP coils are measured by integrating the products of almost resistive components extracted from the terminal voltages in the HTS coil and applied transport currents observed using a pickup coil over a cycle under AC operations. Net AC losses in the two-ply bundle conductor windings are obtained by subtracting the Joule losses from the measured ones. In order to understand the mechanism of AC losses in the two-ply bundle conductor windings, the influences of current amplitudes and frequencies are investigated experimentally and theoretically.

Recognizing Customer Behavior (CB) from videos of instore cameras is important to smart retail solutions. Because of possible changes in retail needs and environments, a high degree of adaptability to different target CBs is required for Customer Behavior Recognition (CBR) methods. Existing CBR methods are mainly machine learning based models due to their remarkable recognition accuracy. However, trained models are not reusable for different target CBs. Consequently, existing CBR methods are hard to adapt to different target CBs because the necessary recollecting data and retraining models. In this paper, we propose a CBR method that recognizes CBs by combinations of primitives, each of which represents an object’s motion or objects’ relationship. Since primitives can be reused in combinations for various CBs, the proposed method is easily adaptable to changed target CBs. Experiments on two datasets indicate the good adaptability and sufficient recognition accuracy of our method.

Local degradation of critical currents in REBa2Cu3Oy (REBCO, RE: rare-earth and Y) coated conductor is one of the serious issues especially for high field superconducting magnets. It may give rise to a localized hotspot, followed by a burn-out in the magnet. To mitigate such phenomena, a co-winding of bundled two REBCO coated conductors is considered. The bundled REBCO double pancake coil with a local damaged area was made and tested in LN2 and conduction cooling conditions. The Ic of the damaged coil showed more than 90 % of that of the non-damaged coil. The comparison of the Ic distribution in the coil and the Ic value at the damaged area can explain the good Ic performance of the damaged coil. Even if the operation current exceeds the local Ic due to the local damage, the coil can be protected if we would detect the quench with a threshold of less than 18 mV for the practical 30 T cryogen-free superconducting magnet.

This paper proposes a Software-Defined Network (SDN)-based Moving Target Defense (MTD) to protect the network from potential scans in a compromised network. As a unique feature, contrary to tradi-tional MTDs, the proposed MTD can work alongside other tools and coun-termeasures already deployed in the network (e.g., Intrusion Protection and Detection Systems) without affecting its behavior. Through extensive eval-uation, we showed the effectiveness of the proposed mechanism compared to existing solutions in preventing scans of different rates without affecting the network and controller performance.

To provide analytic materials for business management for smart retail solutions, it is essential to recognize various customer behaviors (CB) from video footage acquired by in-store cameras. Along with frequent changes in needs and environments, such as promotion plans, product categories, in-store layouts, etc., the targets of customer behavior recognition (CBR) also change frequently. Therefore, one of the requirements of the CBR method is the flexibility to adapt to changes in recognition targets. However, existing approaches, mostly based on machine learning, usually take a great deal of time to re-collect training data and train new models when faced with changing target CBs, reflecting their lack of flexibility. In this paper, we propose a CBR method to achieve flexibility by considering CB in combination with primitives. A primitive is a unit that describes an object's motion or multiple objects' relationships. The combination of different primitives can characterize a particular CB. Since primitives can be reused to define a wide range of different CBs, our proposed method is capable of flexibly adapting to target CB changes in retail stores. In experiments undertaken, we utilized both our collected laboratory dataset and the public MERL dataset. We changed the combination of primitives to cope with the changes in target CBs between different datasets. As a result, our proposed method achieved good flexibility with acceptable recognition accuracy.