In this paper, we discuss a trajectory decision problem of Unmanned Aerial Vehicles (UAVs) for UAV-based edge computing systems. In the UAV-based edge computing systems, some UAVs having computing resources fly over a given area to process data generated by clients such as IoT devices. In order to utilize the computing resources of UAVs, it is important for us to consider the trajectory of UAVs for approaching client nodes. Because UAVs have constraints about moving distance due to the power consumption, we need to consider the placement of charging stations in addition to the UAV trajectory. For this problem, this paper proposes an optimization method of the trajectory of multiple UAVs in a given area, which is a variant of the traveling salesman problem. The proposed method aims to minimize the maximum traveling distance of the UAVs. We divide the problem into some sub-problems such as clustering and trajectory decision to obtain approximate solutions of the optimization problem. Through numerical experiments, we show the effectiveness of the proposed method.

The purpose of this study was to evaluate the precision of various heart rate variability (HRV) analysis parameters during postural changes between sitting and upright postures, with a particular focus on Tone-Entropy analysis, and to verify the absolute evaluability and high reproducibility of Tone-Entropy analysis. Fifteen healthy young men (age 20.7±1.5 years) from two different groups participated in the study, and their heart rates (HRs) were recorded during sitting and upright postures. The postural changes were repeated twice, and each posture was maintained for 2 minutes. Precision was evaluated by defining the error rate for each HRV index. As a result, Entropy showed almost the same precision as HR. On the other hand, standard frequency domain indices such as low frequency (LF), high frequency (HF), and LF/HF ratio showed consistently low precision. Furthermore, our results support the usefulness of Tone-Entropy analysis to date, as it can allow absolute evaluation in Tone-Entropy space even for different groups of subjects and experimental conditions, and the reproducibility of Entropy is also high. In conclusion, it is suggested that Tone-Entropy analysis has high reliability and reproducibility and that frequency power spectrum analysis lacks sufficient precision to be applied to long-term monitoring analyses in young male subjects.

In this study, we developed a computerized classification method for molecular subtypes in LGGs (low-grade glioma) using multi-scale 3D-dual attention branch networks (MS3D-DABNs) with ElasticFace loss. Our database included T1-weighted, T2-weighted, and FLAIR brain MRI images from 217 patients (58 IDH-mutant astrocytomas, 49 IDH-wildtype astrocytomas, and 110 oligodendrogliomas) from Tokyo Women's Medical University. The proposed network consisted of a feature extractor, a dual attention branch combining spatial and channel attention modules, and a perception branch. The feature extractor first extracted the feature maps at different resolutions from each MRI image, which were then sequentially processed by the spatial and channel attention modules. The spatial attention module weighted features on the tumor region in the feature maps, while the channel attention module weighted important features in the feature maps. Finally, the perception branch classified LGGs into three molecular subtypes. During the training of the proposed network, ElasticFace loss was used to optimize intra-class compactness and inter-class separability. The average classification accuracy of the proposed method was 72.8% (Astrocytoma IDH-mutant: 65.5%, Astrocytoma IDH-wildtype: 71.4%, and Oligodendroglioma: 77.2%), showing an improvement compared to our previous network, multi-scale 3D spatial attention branch networks with ArcFace loss, 66.4% (51.7%, 61.2%, and 76.4%).

Sleep apnea syndrome (SAS) is one of the most common sleep-related diseases. SAS can reduce sleep quality and is frequently associated with lifestyle-related diseases. Many individuals are unaware that they have this disease as SAS occurs during sleep. Although polysomnography (PSG) is used to diagnose SAS, it is not suitable for sleep monitoring on a daily basis as it requires hospitalization and the use of expensive wearable devices. The authors previously proposed a method for monitoring breathing during sleep using a depth camera in a non-contact and non-constrained manner. However, the time spent by the subject in the correct position could not be estimated because objects moving within the camera's field of view could be erroneously identified as persons. In this study, we attempt to separate beds from other room areas and identify multiple person-like moving objects within the beds using distance information from a depth camera. Consequently, we show that it is possible to estimate a person's position and respiration without interference from the surrounding environment and that the accuracy of these estimations can be improved with respect to that of the conventional method.

To assist people with severe disabilities to eat and drink at their own pace, we present a self-feeding robotic arm system with the functions of object detections using YOLOv5 model and user's mouth detection using MediaPipe. The redundant robotic arm with 7-DOF (degree-of-freedom) was made to perform natural eating motion and avoid obstacles by considering the home care and nursing care environments in Japan. An open-source software, ROS (Robot Operating System), and a motion planning framework, MoveIt! simulator including the motion planning solver using inverse kinematics, were used to simulate the 7-DOF robotic arm. Moreover, a single-finger operated interface was applied as a controller for the robotic arm system. We demonstrated that using the proposed robotic arm system, the tasks to grasp and move the plastic bottle to the user's mouth were conducted. From the simulation and experimental results, we found that the detections of the target object and the user's mouth were conducted effectively. In addition, it was shown that the success rates of the tasks were more than 80% or equal when having no object in the calculated trajectory of the robotic arm. Future works include conducting some experiments of the tasks with people with disabilities.

On January 1, 2024, at 16:10 JST, a magnitude (Mj) 7.6 earthquake struck the Noto Peninsula in Ishikawa Prefecture. The earthquake caused extensive damage in Ishikawa Prefecture, with more than 300 people killed and over 6,000 buildings damaged (as of September 2024). In the Noto Peninsula earthquake, several organizations are attempting to use satellites and other means to obtain an overview of the earthquake damage from the sky in order to quickly assess the extent of the damage after the earthquake. Here, we examine the “building damage potential” data from this study, which was conducted using satellite-based SAR imagery. The study organized the relationship between “building damage potential” based on SAR images and measured seismic intensity, a typical seismic intensity index. Comparisons with actual building damage were made to confirm the validity of the estimation results using Multi-wavelength SAR.

In this paper, we consider large-scale mean-field stochastic systems. After defining a stabilization problem based on a static output feedback strategy, we apply Nash equilibrium strategies to the problem. Note that the initial condition assumption is extended to the general case compared to existing results. The problem of minimizing the cost function is solved using the Lagrange multiplier technique. However, as the number of players increases, computational challenges arise in determining solutions. To address this problem, we develop a decentralized design and the numerical algorithm to obtain the suboptimal solution set. As a result, the parameter-independent strategy and the associated function are investigated. The Pareto strategy is also considered as another cooperative game. Numerical examples validate the practicality and effectiveness of the decentralized numerical algorithm.

This study addresses the control challenges involved in automating an excavator's operation for loading soil onto a dump truck. For an efficient loading motion, the tip path should be minimized to avoid contacting the truck. It is also crucial to prevent soil from spilling out of the bucket during the loading process. Model Predictive Control (MPC) is well-suited for managing hydraulic excavators under these multiple performance criteria. However, obtaining accurate models of hydraulic actuator dynamics is challenging because they vary significantly with factors such as oil temperature and excavator posture. When accurate models are not available, MPC may not deliver adequate control performance. To address this issue, we propose a method that compensates for discrepancies between the actual actuator dynamics and the internal model assumed in MPC by integrating actuator compensators. These compensators are tuned in real time using a data-driven control approach. An additional advantage of this method is that the MPC, which is tuned based on a nominal model, continues to perform effectively even when the actuator dynamics change. The effectiveness of the proposed method was validated through numerical simulations.

This research proposes a 2D mapping method that considers the potential occupancy space of mobile objects for a guide dog robot, aiming at enhancing safe navigation for visually impaired individuals. By using a RS-LiDAR-16 sensor and an RGB-D camera, the guide dog robot realized robust 2D mapping by integrating cartographer based 2D mapping, object recognition results by using YOLACT, and Harris Corner detection. Meanwhile, a novel dynamic risk map, responsive to the robot’s position, is developed to avoid collisions with suddenly obstacles appearing from blink spots like doors and corners. Experimental results demonstrate that the generated dynamic risk map significantly improves the performance of collision avoidance, reduces moving time, and increases the flexibility of the guidance route, proving the effectiveness of the proposed method in dynamic environments.

Unpaired image-to-image translation is a task that is expected to be applied in scientific simulations and other applications. In particular, the conversion of selfie photos to animated images is expected to be applied in the production of animation, manga, games, etc. CycleGAN was the first model to achieve unpaired image-to-image translation tasks, and many methods have been proposed to apply it since then. However, these methods often had difficulty with geometric transformations, and it was difficult to achieve selfie photo to animation translations involving large geometric transformations for the eyes and nose. In this study, we focus on a CycleGAN-based method that can perform the transformation with an intuitive and simple architecture, and introduce a mechanism to assist geometric transformation into UVCGAN, which introduces the Vision Transformer to CycleGAN. The goal is to improve the performance of the conversion from selfie photos to animated images by generating natural-looking images. For this purpose, landmark images of the original and reconstructed images are obtained when calculating the cycle-consistency loss, and the difference between them is also accounted for in the loss function. This method produces more natural-looking images than the conventional method, as the eye sizes and facial contours are now aligned.

IIRC Framework, which provides guidelines for integrated reports, states that the purpose of an integrated report is to explain the process by which corporate value is formed in the long term. Based on this framework, the purpose of this study is to reveal the relationship between integrated reports prepared in accordance with the IIRC framework and stock prices. This study focuses on integrated reports issued by Japanese companies. First, betweenness and degree centralities are used to score the integrated reports prepared according to the IIRC framework. Next, portfolios are created by quantile according to the score, and regression analysis is performed using Fama-French three factor model. As a result, a positive abnormal return is observed in the portfolio with the highest score stocks. It indicates the potential for this score to be used in stock selection.