Hak-Sub Kim

Saliency Prediction on Stereoscopic Videos

We describe a new 3D saliency prediction model that accounts for diverse low-level luminance, chrominance, motion, and depth attributes of 3D videos as well as high-level classifications of scenes by type. The model also accounts for perceptual factors, such as the nonuniform resolution of the human eye, stereoscopic limits imposed by Panums fusional area, and the predicted degree of (dis) comfort felt, when viewing the 3D video. The high-level analysis involves classification of each 3D video scene by type with regard to estimated camera motion and the motions of objects in the videos. Decisions regarding the relative saliency of objects or regions are supported by data obtained through a series of eye-tracking experiments. The algorithm developed from the model elements operates by finding and segmenting salient 3D space-time regions in a video, then calculating the saliency strength of each segment using measured attributes of motion, disparity, texture, and the predicted degree of visual discomfort experienced. The saliency energy of both segmented objects and frames are weighted using models of human foveation and Panums fusional area yielding a single predictor of 3D saliency.

Globular cluster systems of early-type galaxies in low-density environments

Deep images of 10 early-type galaxies in low-density environments have been obtained with the Advanced Camera for Surveys (ACS) aboard the Hubble Space Telescope. The global properties of the globular cluster (GC) systems of the galaxies have been derived in order to investigate the role of the environment in galaxy formation and evolution. Using the ACS Virgo Cluster Survey as a high-density counterpart, the similarities and differences between the GC properties in high- and low-density environments are presented. We find a strong correlation of the GC mean colours and the degree of colour bimodality with the host galaxy luminosity in low-density environments, in good agreement with high-density environments. In contrast, the GC mean colours at a given host luminosity are somewhat bluer [Δ(g−z) ∼ 0.05] than those for cluster galaxies, indicating more metal poor (Δ[Fe/H] ∼ 0.10 − 0.15) and/or younger (Δage > 2 Gyr) GC systems than those in dense environments. Furthermore, with decreasing host luminosity, the colour bimodality disappears faster, when compared to galaxies in cluster environments. Our results suggest that: (1) in both high- and low-density environments, the mass of the host galaxy has the dominant effect on GC system properties; (2) the local environment has only a secondary effect on the history of GC system formation; and (3) GC formation must be governed by common physical processes across a range of environments.

Transition of Visual Attention Assessment in Stereoscopic Images With Evaluation of Subjective Visual Quality and Discomfort

Through statistical analysis of the behaviors of human visual attention, we discovered that fixation behaviors are highly correlated with the degree to which the viewer experiences visual quality/discomfort with stereoscopic images. To quantify the correlation between visual attention and quality/discomfort with respect to various distortions and disparities, we explore a novel methodology called transition of visual attention (ToVA) that accounts for diverse low-level luminance, chrominance, and depth attributes of 3D images in eye-tracker experiments. Moreover, saliency entropy is defined in order to quantify the distribution of fixations for 3D images. Using saliency entropy, we interpret perceptual factors in accordance with the non-uniform resolution of the human eye and the stereoscopic limits caused by foveation and Panums fusional area. We then measure ToVA in terms of relative saliency entropy using Kullback-Leibler divergence. To evaluate the effectiveness of ToVA, a successful example application is also provided for which ToVA is applied to obtain subjective measurements of the quality and discomfort experienced when viewing 3D displays, rather than relying on conventional subjective tests using a scoring system.

A cross-layer optimization for energy-efficient MAC protocol with delay and rate constraints

We propose the energy efficient MAC algorithm in this paper. In the proposed algorithm, each node sets the contention window size with respect to the residual energy, the harvesting energy and the transmit power. This algorithm makes the sensor nodes consume their energy efficiently. To achieve this goal, we use the game theory and the cross-layer optimization. Introducing the non-cooperative game, we can formulate the utility function easily. In this paper, we can allocate the optimal power by the cross-layer optimization on the PHY and the MAC.

Implementation of DWT-based adaptive mode selection for LCD overdrive

Liquid crystal display (LCD) overdrive is used to remove motion blur. To effectively realize overdrive, it is necessary to efficiently compress and store previous frames. In this paper, we propose an adaptive mode selection algorithm via discrete wavelet transform (DWT) for efficient image compression while maintaining high visual quality. Moreover, we propose a mode reduction algorithm to adjust the target rate and to easily implement the H/W architecture. We simulated the performance of the proposed method at a fixed target rate and found that its performance was much better than that of conventional schemes.

Implementation of Human Action Recognition System Using Multiple Kinect Sensors

Human action recognition is an important research topic that has many potential applications such as video surveillance, human-computer interaction and virtual reality combat training. However, many researches of human action recognition have been performed in single camera system, and has low performance due to vulnerability to partial occlusion. In this paper, we propose a human action recognition system using multiple Kinect sensors to overcome the limitation of conventional single camera based human action recognition system. To test feasibility of the proposed system, we use the snapshot and temporal features which are extracted from three-dimensional (3D) skeleton data sequences, and apply the support vector machine (SVM) for classification of human action. The experiment results demonstrate the feasibility of the proposed system.

IRAC Observations of CO J = 4 → 3 High-Velocity Cloud in the 30 Doradus Complex in the Large Magellanic Cloud

We present the results of 12CO J = 2 → 1 observations of the X-ray-bright giant shell complex 30 Doradus in the Large Magellanic Cloud (LMC) using the Antarctic Submillimeter Telescope and Remote Observatory (AST/RO). This is the one of the largest H II complexes in the Local Group. We compare the 12CO J = 2 → 1 observations against previously made 12CO J = 4 → 3 observations and analyze the spatial distribution of young stellar objects (YSOs) within the cloud using the Spitzer IRAC observations of the 30 Doradus complex. Both peaks of 12CO J = 2 → 1 and J = 4 → 3 emitting clouds coincide with the densest region of the filaments in which multiple shells are colliding. The YSOs are clustered in the southern ridge of the warm and dense molecular gas clouds traced by 12CO J = 4 → 3, indicating a filamentary structure of star formation throughout 30 Doradus. We also find an excess of Class I YSO candidates close to the clouds, which likely represent the most recent phase of star formation in this region. This is a region where the triggered star formation has actually occurred, and newly formed stars may have produced such a high-velocity outflow through interacting with the surrounding molecular cloud material.

Real-time video traffic management for a warship combat system

In an effort to improve the channel utilization of a warship combat system (WCS), it is necessary to control and manage the large volume of video traffic by considering a strong self-similarity characteristic. We propose a new video traffic management method for the WCS that smooths the traffic of the closed-circuit television surveillance system (CSS) video and encodes the electro-optical tracking system (EOTS) video with low complexity in real time. First, to achieve global and local traffic smoothing for the CSS video traffic, a proper bandwidth allocation interval and a minimum bandwidth for the CSS video traffic must be determined. Second, due to the increased available bandwidth after traffic smoothing, more elaborate low-delay encoding is accomplished by selecting encoding modes adaptively to compress the EOTS video efficiently with high visual quality. The experimental results demonstrate that the proposed method is able to control the WCS traffic to be as flat as possible and to encode the EOTS video with low delay, even in combat and emergency scenarios. Hence, we show that to migrate to network-centric warfare successfully, the proposed method is much more suitable for the WCS than conventional video traffic management.

Discovery of the Kinematic Alignment of Early-type Galaxies in the Virgo Cluster

Using the kinematic position angles (PA_kin), an accurate indicator for the spin axis of a galaxy, obtained from the ATLAS3D integral-field-unit (IFU) spectroscopic data, we discovered that 57 Virgo early-type galaxies tend to prefer the specific PA_kin values of 20 degree and 100 degree, suggesting that they are kinematically aligned with each other. These kinematic alignment angles are further associated with the directions of the two distinct axes of the Virgo cluster extending east-west and north-south, strongly suggesting that the two distinct axes are the filamentary structures within the cluster as a trace of infall patterns of galaxies. Given that the spin axis of a massive early-type galaxy does not change easily even in clusters from the hydrodynamic simulations, Virgo early-type galaxies are likely to fall into the cluster along the filamentary structures while maintaining their angular momentum. This implies that many early-type galaxies in clusters are formed in filaments via major mergers before subsequently falling into the cluster. Investigating the kinematic alignment in other clusters will allow us to understand the formation of galaxy clusters and early-type galaxies.

Visual attention analysis on stereoscopic images for subjective discomfort evaluation

By analyzing the statistical behaviors on human visual attention, we discover a clue that the fixation behaviors are highly correlated with how much the viewers feel visual discomfort on stereoscopic images differently from conventional subjective assessments. In order to quantify the correlation between visual attention and discomfort, we explore a novel methodology termed transition of visual attention (ToVA) according to various disparities, which accounts depth attributes of 3D images by eye-tracker experiments. Moreover, the saliency entropy is defined to quantify the distribution of fixations for 3D images. Then, we measure ToVA in terms of the relative saliency entropy using Kullback-Leibler divergence. In order to evaluate the effectiveness of ToVA, a successful example application is also provided, whereby ToVA is applied to obtaining subjective results of measuring discomfort experienced when viewing 3D displays rather than relying on the conventional subjective test by using scoring system.