Seokjun Hong’s research while affiliated with Korea University and other places

One popular application of augmented reality (AR) is the real-time guidance and training in which the AR user receives useful information by a remote expert. For relatively fast-paced tasks, presentation of such guidance in a way that the recipient can make immediate recognition and quick understanding can be an especially challenging problem. In this paper, we present an AR-based tele-coaching system applied to the game of tennis, called the AR coach, and explore for interface design guidelines through a user study. We have evaluated the player’s performance for instruction understanding when the coaching instruction was presented in four different modalities: (1) Visual—visual only, (2) Sound—aural only/mono, (3) 3D Sound—aural only/3D and (4) Multimodal—both visual and aural/mono. Results from the experiment suggested that, among the three, the visual-only augmentation was the most effective and least distracting for the given pace of information transfer (e.g., under every 3 s). We attribute such a result to the characteristic of the visual modality to encode and present a lot of information at once and the human’s limited capability in handling and fusing multimodal information at a relatively fast rate.

In this paper, we present a case study of designing and improving an interface for a web based security analysis software for the internet of things (IoT), called the IoTCube. The objective of the IoTCube is to provide an easy-to-use security vulnerability checking and analysis solution for the IoT related developers and users. The software is consisted and organized of largely three analysis functionalities for the: (1) system/executables, (2) source code and (3) network protocols. The interface design started with the user analysis and deriving of requirements based on usage scenarios, and then went through several iterations of sketches, wire-frames and implementations based on feedbacks from the HCI experts, domain experts and actual users to further improve its usability. The improved usability not only has made the security analysis practice much accessible for the non-experts, but also brought about a concrete understanding of the consequences of identified problems.

In this paper, we present a functionality of resetting the viewpoint into the 360 degree video relative to the user's body pose. Such a functionality can improve the usability of the 360 degree videos in terms of reducing excessive movement, thereby reducing fatigue and more usable in the public setting. Two variant methods are proposed and compared to the baseline case (no resetting possible) in terms of usability and content understanding. In both methods, the user can select a particular direction and associated view, then turn the body to a nominal direction and apply the reset. In method 1, the video stops (screen going to black momentarily) in between these actions, while in 2, the video continues. The Method 1 was found to be the viable method for viewpoint recalibration for casual viewing in a constrained public setting that can minimize excessive physical movement, yet not much different from the baseline case (freely rotate in 360 degrees) in terms of sickness, disorientation and content understandability.

In this paper, we present an application of rotational gain to horizontal panning for viewing 360 degree videos. Rotational gain refers to the ratio between the rotation velocity in the virtual (video) space to that of the physical, and it allows the user to rotate one's head less quickly than actually needed (without the user noticing such adjustment to some degree). As such it can bring about convenience, and less physical movement, fatigue and possibly even sickness. We implemented a 360 video panning system with both a constant and dynamic gain, and compared user behavior and subjective usability. Our pilot study showed promising results in that with the proper gain value and control scheme, the user will unknowingly use less physical movement than needed, yet maintain reasonable spatial understanding with higher usability.

In this paper, we present an AR-based tele-coaching system for fast-paced tasks, applied to the game of tennis, and explore for interface design guidelines through a user study. We have evaluated the player's performance for instruction understanding in three different modalities of augmentation: (1) AR --- visual only, (2) Sound --- aural only and (3) Multimodal --- both visual and aural. The augmented instructions were useful even in the stringent temporal conditions, and most effective with the visual only augmentation due to its modal capability of encoding and presenting several information compactly at once.