Inwoong Lee

A Pervasive Network Control Algorithm for Multicamera Networks

Owing to the increasingly large volume and complexity of captured videos, renewable energy systems based on solar energy are of particular interest in the design of energy harvesting (EH) wireless visual sensor networks (WVSNs). Since additional energy consumption following image capture occurs owing to image processing, mote operation, data transmission, and reception, the capture rate significantly affects the lifetime of a node. To this end, we explore a novel energy-efficient framework for EH-WVSN design by developing an optimal algorithm named capture rate and pervasive network control for multicamera networks where the quality of service is maximized by obtaining optimal values for the capture rate, allocated energy, and transmit power, based on field of view-based networking in the presence of event and power acquisition patterns. Through simulations, we demonstrate the feasibility of EH-WVSNs in terms of energy consumption, energy allocation, and capture rate in a realistic scenario (parking surveillance).

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.

Theoretical-Analysis-Based Distributed Load Balancing Over Dynamic Overlay Clustering

In multicell networks, unbalanced cell loading can lead to decreased system stability and reduced fairness among serviced users. In this paper, we propose theoretical-analysis-based distributed load balancing (DLB) over dynamic overlay clustering implemented over a multicell network. The proposed system is divided into two parts: DLB and overlay clustering. First, for DLB, we define the long-term expected load after deriving the long-term expected rate in terms of proportional fairness. We then introduce two algorithms: DLB for load dispersion and DLB for edge-rate enhancement (ERE). These algorithms operate in a distributed manner based on mathematical analyses and load balancing characteristics. Second, through overlay clustering, load balancing within each cluster is consecutively performed on neighboring clusters, which enables the algorithm to optimally approximate in a distributed manner. The simulation results show that approximately 90% of the near-optimal performance in terms of load variation and ERE can be achieved with low complexity by using the proposed schemes. In addition, we discuss aspects and tradeoffs of the load balancing system.

Propagating LSTM: 3D Pose Estimation based on Joint Interdependency

We present a novel 3D pose estimation method based on joint interdependency (JI) for acquiring 3D joints from the human pose of an RGB image. The JI incorporates the body part based structural connectivity of joints to learn the high spatial correlation of human posture on our method. Towards this goal, we propose a new long short-term memory (LSTM)-based deep learning architecture named propagating LSTM networks (p-LSTMs), where each LSTM is connected sequentially to reconstruct 3D depth from the centroid to edge joints through learning the intrinsic JI. In the first LSTM, the seed joints of 3D pose are created and reconstructed into the whole-body joints through the connected LSTMs. Utilizing the p-LSTMs, we achieve the higher accuracy of about 11.2% than state-of-the-art methods on the largest publicly available database. Importantly, we demonstrate that the JI drastically reduces the structural errors at body edges, thereby leads to a significant improvement.

Optimal Beam Steering for Maximal Visual Quality Over a Multimedia Broadcasting System

Various types of multimedia services have become prominent over the last few years leading to difficulty in providing high-quality broadcast multimedia services in certain shadowing regions without resource management in accordance with specific channel adaptation. In order to overcome this difficulty, return channel design has been actively discussed in broadcasting standards for broadcasting systems. Context-aware multimedia services have been developed by adapting radio resources according to service type. In order to forward this goal, we propose a beamforming technique to optimize spatial resources in order to maximize visual quality for video broadcasting services. In order to enhance the video broadcast service, we derive expected visual information (EVI) that quantifies the visual quality of video and adapt the EVI into multiple-input and multiple-output beamforming techniques in a video broadcast system. In order to guarantee the proportional fairness of broadcast users, we propose a novel beam steering algorithm using the log-EVI sum maximization, and compare it to the traditional channel-quality-based max-min problem. The beam-steering algorithm is performed iteratively while switching the bases, which constructs a beam steering weight vector domain. In the simulation, we demonstrate that the proposed algorithm has less computational complexity than conventional algorithms, and the visual quality of the proposed algorithm is enhanced over that of the traditional quasi-optimal max-min fairness algorithm.

Optimal phase control for joint transmission and reception with beamforming

In this paper, we propose a joint transmission and reception with phase control for beamforming in a multi-cell environment. For generated transmit weight vectors of multiple base stations (BSs), a mobile station (MS) calculates phases to maximize an achievable rate with low rate feedback. By using the phases, the multiple transmit weight vectors are coordinated to improve the signal to noise ratio (SNR). In order to find optimal phases, we present a phase control method for the effective channel via geometrical approach.

Combinatorial JPT based on orthogonal beamforming for two-cell cooperation

In this paper, we investigate efficient multi-cell cooperation based on CoMP-joint processing and transmission (CoMP-JPT) with orthogonal beamforming. Through the use of a combinatorial optimization algorithm, the optimal user scheduling for joint transmission using multiple transmitters is accomplished. The throughput of the CoMP-JPT can be significantly improved while maintaining fairness among users over a multi-cell environment.