In Augmented Reality (AR), search performance for outdoor tasks is an important metric for evaluating the success of a large number of AR applications. Users must be able to find content quickly, labels and indicators must not be invasive but still clearly noticeable, and the user interface should maximize search performance in a variety of conditions. To address these issues, we have set up a series of experiments to test the influence of virtual characteristics such as color, size, and leader lines on the performance of search tasks and noticeability in both real and simulated environments. The first experiment showed that limited FOV will severe-ly limit search performance, but that appropriate placement of labels and leaders within the periphery can alleviate this problem without interfering with walking or decreasing user comfort. In the second experiment, we found that different types of motion are more no-ticeable in optical versus video see-through displays, but that blue coloration is most noticeable in both. Results can aid in designing more effective view management techniques, especially for wider field of view display.

In this article, a gait recognition algorithm is presented based on the information obtained from inertial sensors embedded in a smartphone, in particular, the accelerometers and gyroscopes typically embedded on them. The algorithm processes the signal by extracting gait cycles, which are then fed into a Recurrent Neural Network (RNN) to generate feature vectors. To optimize the accuracy of this algorithm, we apply a random grid hyperparameter selection process followed by a hand-tuning method to reach the final hyperparameter configuration. The different configurations are tested on a public database with 744 users and compared with other algorithms that were previously tested on the same database. After reaching the best-performing configuration for our algorithm, we obtain an equal error rate (EER) of 11.48% when training with only 20% of the users. Even better, when using 70% of the users for training, that value drops to 7.55%. The system manages to improve on state-of-the-art methods, but we believe the algorithm could reach a significantly better performance if it was trained with more visits per user. With a large enough database with several visits per user, the algorithm could improve substantially.

To create rich experiences in virtual reality (VR) environments, it is essential to define the behavior of virtual objects through programming. However, programming in 3D spaces requires a wide range of background knowledge and programming skills. Although Large Language Models (LLMs) have provided programming support, they are still primarily aimed at programmers. In metaverse platforms, where many users inhabit VR spaces, most users are unfamiliar with programming, making it difficult for them to modify the behavior of objects in the VR environment easily. Existing LLM-based script generation methods for VR spaces require multiple lengthy iterations to implement the desired behaviors and are difficult to integrate into the operation of metaverse platforms. To address this issue, we propose a tool that generates behaviors for objects in VR spaces from natural language within Cluster, a metaverse platform with a large user base. By integrating LLMs with the Cluster Script provided by this platform, we enable users with limited programming experience to define object behaviors within the platform freely. We have also integrated our tool into a commercial metaverse platform and are conducting online experiments with 63 general users of the platform. The experiments show that even users with no programming background can successfully generate behaviors for objects in VR spaces, resulting in a highly satisfying system. Our research contributes to democratizing VR content creation by enabling non-programmers to design dynamic behaviors for virtual objects in metaverse platforms.

This paper presents a deep data-driven inertial measurement unit (IMU) online calibration (DUET) method that can compensate for the run-time errors of the accelerometer and gyroscope to improve inertial-based odometry. We design a differential error learning strategy based on the kinematic motion model to train the sensor error compensation model. This strategy allows our method to learn IMU sensor errors, such as scale factors, axis-misalignment, and biases, solely from displacement and orientation increments given by external tracking systems. Then during the odometry computation, the trained model leverages the past inertial data to mitigate the sensor errors and thus reduces the integration errors to reflect the odometry state. The experiments conducted on two public visual-inertial datasets show an average of 20% improvement in the position estimation accuracy of visual-inertial odometry, which is comparable to existing learning-based methods with lower operational complexity.

Over time, the desks of office workers often get cluttered with documents, items, and other objects that are unrelated to the work being done. These objects become visual noise. Visual noise tends to reduce the worker’s concentration on their work. Therefore, it is desirable to have an uncluttered environment to work. We propose DecluttAR, a system that controls the visibility of clutter objects by using a video see-through head-mounted display and an RGB-D camera. Our system first generates a 3D model of the empty desk, and then automatically generates a 3D model of each object that is newly placed on the desk. Then, the user can interactively control each object’s visibility which is superimposed on the corresponding real object. Four types of object appearance adjustments are implemented: normal, grayscale, outline, and transparency. To investigate changes in the ability to concentrate in different appearances, a user study was conducted in which a simple arithmetic task was performed in four appearances. The results show that the number of answers in the arithmetic task for the transparent display increased compared to the normal display in the first half of the task. This result indicates that the participants are able to enter a state of concentration on the task more quickly by making the object invisible. In addition, the questionnaire-based subjective evaluation revealed that the transparent display had a lower workload and a higher user evaluation followed by the outline display.

We present computational phase-modulated eyeglasses, a see-through optical system that modulates the view of the user using phase-only spatial light modulators (PSLM). A PSLM is a programmable reflective device that can selectively retardate, or delay, the incoming light rays. As a result, a PSLM works as a computational dynamic lens device. We demonstrate our computational phase-modulated eyeglasses with either a single PSLM or dual PSLMs and show that the concept can realize various optical operations including focus correction, bi-focus, image shift, and field of view manipulation, namely optical zoom. Compared to other programmable optics, computational phase-modulated eyeglasses have the advantage in terms of its versatility. In addition, we also presents some prototypical focus-loop applications where the lens is dynamically optimized based on distances of objects observed by a scene camera. We further discuss the implementation, applications but also discuss limitations of the current prototypes and remaining issues that need to be addressed in future research.