Kwangyun Wohn

Recognition of hand gestures with 3D, nonlinear arm movement

Abstract Hand gesture is a useful modality of human interaction. In this paper, we propose an approach to recognition of space-time variable patterns of nonlinear arm movement and integration with other attributes to find the proper interpretation. At the encoding stage, we first extract the essential 2D trajectory from 3D arm movement by a plane fitting method. Pause information between the consecutive gestures is also modeled and integrated into the encoding. Codified information is then applied is a hidden Markov model (HMM) network which is responsible for segmentation and recognition of continuous arm movements. As a whole, three major attributes of hand gestures are processed in parallel and independently, followed by the inter-attribute communication for finding the proper interpretation.

Concurrency control in CIAO

This paper is concerned with the concurrency control for collaborative virtual environments. In particular, we describe how concurrent actions are coordinated in a multi-user, large-scale 3-D layout system CIAO. In contrast to many existing systems that sacrifice responsiveness in order to maintain consistency, CIAO achieves optimal response and notification time without compromising consistency. The optimal responsiveness is achieved by a new multicast-based, optimistic concurrency control mechanism. Even operations on a group of related objects do not entail any latency for concurrency control. We also present the multi-user interfaces of CIAO that provide some sense of isolation as well as rich awareness.

The control of avatar motion using hand gesture

It is difficult to navigate virtual environment as in real world and to interact with other participant in virtual environment, especially wearing Head-Mounted Display (HMD). We developed Virtual Office Environment System (VOES), and avatar is used to navigate and to interact with other participants. For easy and intuitive control of avatar motion in the system, we use continuous hand gesture recognition system. State automata are proposed in hand gesture recognition to segment continuous hand gesture and to remove meaningless motion. Using avatar and gesture interface, this system provides natural navigation and interaction in virtual environment system.

MagGetz: customizable passive tangible controllers on and around conventional mobile devices

This paper proposes user-customizable passive control widgets, called MagGetz, which enable tangible interaction on and around mobile devices without requiring power or wireless connections. This is achieved by tracking and ana-lyzing the magnetic field generated by controllers attached on and around the device through a single magnetometer, which is commonly integrated in smartphones today. The proposed method provides users with a broader interaction area, customizable input layouts, richer physical clues, and higher input expressiveness without the need for hardware modifications. We have presented a software toolkit and several applications using MagGetz.

KAIST interactive bicycle simulator

This paper presents key technologies and system integration issues of the KAIST interactive bicycle simulator. The rider on the bicycle feels the motion and has the visual experience as if he/she is riding in the campus of the Korea Advanced Institute of Science and Technology. The simulator consists of a bicycle, a Stewart platform, a magnetorheological handle, a pedal resistance system to generate motion feelings, a real-time visual simulator and projection system, sub-controllers and an integrating control network.

Low damped cloth simulation

This paper proposes a new implicit integration technique that reproduces a stable cloth without introducing excessive damping forces. Semi-implicit integration methods have been widely used in cloth simulations because of their high stability and speed. Artificial internal damping forces are generated during the linearization process of the semi-implicit integration. The simulations become extremely stable due to the artificial forces; however, the forces significantly degrade the realism of cloth simulations, since they are generated with respect to rotational rigid motions, as well as internal deformations. Hence, we propose a new method to decrease the damping artifacts. The artificial internal damping forces are computed solely for pure internal deformations, and are stably incorporated into the dynamical system. Experiments show that our simulator can reproduce various cloth materials without excessive damping artifacts even in real-time.

Tracking and Motion Estimation of the Articulated Object

Real-time motion capture plays a very important role in various applications, such as 3D interface for virtual reality systems, digital puppetry, and real-time character animation. In this paper we challenge the problem of estimating and recognizing the motion of articulated objects using theoptical motion capturetechnique. In addition, we present an effective method to control the articulated human figure in realtime.The heart of this problem is the estimation of 3D motion and posture of an articulated, volumetric object using feature points from a sequence of multiple perspective views. Under some moderate assumptions such as smooth motion and known initial posture, we develop a model-based technique for the recovery of the 3D location and motion of a rigid object using a variation of Kalman filter. The posture of the 3D volumatric model is updated by the 2D image flow of the feature points for all views. Two novel concepts � the hierarchical Kalman filter (KHF) and the adaptive hierarchical structure (AHS) incorporating the kinematic properties of the articulated object � are proposed to extend our formulation for the rigid object to the articulated one. Our formulation also allows us to avoid two classic problems in 3D tracking: the multi-view correspondence problem, and the occlusion problem. By adding more cameras and placing them appropriately, our approach can deal with the motion of the object in a very wide area. Furthermore, multiple objects can be handled by managing multiple AHSs and processing multiple HKFs.We show the validity of our approach using the synthetic data acquired simultaneously from the multiple virtual camera in a virtual environment (VE) and real data derived from a moving light display with walking motion. The results confirm that the model-based algorithm works well on the tracking of multiple rigid objects.

VibPress: estimating pressure input using vibration absorption on mobile devices

This paper introduces VibPress, a software technique that enables pressure input interaction on mobile devices by measuring the level of vibration absorption with the built-in accelerometer when the device is in contact with a damping surface (e.g., users hands). This is achieved using a real-time estimation algorithm running on the device. Through a user evaluation, we provide evidence that this system is faster than previous software-based approaches, and accurate as hardware-augmented approaches (up to 99.7% accuracy). With this work, we also provide an insight about the maximum number of pressure levels that users can reliably distinguish, reporting usability metrics (time, errors and cognitive load) for different pressure levels and types of gripping gestures (press and squeeze).

Realtime collision detection for virtual reality applications

Virtual reality technology aims at the expansion of the communication bandwidth by providing users with 3D immersive environments. For the true direct manipulation of the environments, fast collision detection must be provided to increase the sense of reality. A collision detection scheme for virtual reality applications is proposed. The method exploits a hierarchical object representation to facilitate the detection of colliding segments.<<ETX>>

MagPen: magnetically driven pen interactions on and around conventional smartphones

This paper introduces MagPen, a magnetically driven pen interface that works both on and around mobile devices. The proposed device is accompanied by a new vocabulary of gestures and techniques that increase the expressiveness of the standard capacitive stylus. These techniques are: 1) detecting the orientation that the stylus is pointing to, 2) selecting colors using locations beyond screen boundaries, 3) recognizing different spinning gestures associated with different actions, 4) inferring the pressure being applied to the pen, and 5) identifying various pens associated with different operational modes. These techniques are achieved using commonly available smartphones that sense and analyze the magnetic field produced by a permanent magnet embedded in a standard capacitive stylus. This paper explores how magnets can be used to expand the design space of current pen interaction, and proposes a new technology to achieve such results.