Bong-Kee Sin

Hand gesture recognition based on dynamic Bayesian network framework

In this paper, we propose a new method for recognizing hand gestures in a continuous video stream using a dynamic Bayesian network or DBN model. The proposed method of DBN-based inference is preceded by steps of skin extraction and modelling, and motion tracking. Then we develop a gesture model for one- or two-hand gestures. They are used to define a cyclic gesture network for modeling continuous gesture stream. We have also developed a DP-based real-time decoding algorithm for continuous gesture recognition. In our experiments with 10 isolated gestures, we obtained a recognition rate upwards of 99.59% with cross validation. In the case of recognizing continuous stream of gestures, it recorded 84% with the precision of 80.77% for the spotted gestures. The proposed DBN-based hand gesture model and the design of a gesture network model are believed to have a strong potential for successful applications to other related problems such as sign language recognition although it is a bit more complicated requiring analysis of hand shapes.

Locating characters in scene images using frequency features

This paper presents a (language-independent) method of locating rectangular text regions in natural scene images. The method consists of two steps that can be applied in succession or independently: the frequency of edge pixels across vertical and horizontal scan lines, and the fundamental frequency in the Fourier domain. The frequency feature of text images is highly intuitive, and this is the focus of the research. The detection of rectangles using a Hough transform is also addressed. Texts that are meaningful to many viewers usually appear in rectangles of colours of high contrast to the background. Hence it is natural to assume that the detection of rectangles may be helpful for locating desired texts correctly in natural outdoor scene images.

Recognizing hand gestures using dynamic Bayesian network

In this paper, we describe a dynamic Bayesian network or DBN based approach to both two-hand gestures and one-hand gestures. Unlike wired glove-based approaches, the success of camera-based methods depends greatly on image processing and feature extraction results. So the proposed method of DBN-based inference is preceded by fail-safe steps of motion tracking. Then a new gesture recognition model for a set of both one-hand and two-hand gestures is proposed based on the dynamic Bayesian network framework which makes it easy to represent the relationship among features and incorporate new information to the model. In an experiment with ten isolated gestures, we obtained a recognition rate upwards of 99.59% with cross validation. The proposed model is believed to have a strong potential for successful applications to other related problems such as sign languages.

HMM-Based gait recognition with human profiles

Recently human gait has been considered as a useful biometric supporting high performance human identification systems. We propose a view-based pedestrian identification method using the dynamic silhouettes of a human body modeled with the hidden Markov model (HMM). Two types of gait models have been developed both with a cyclic architecture: one is a discrete HMM method using a self-organizing map-based VQ codebook and the other is a continuous HMM method using feature vectors transformed into a PCA space. Experimental results showed a consistent performance trend over a range of model’s parameters and the recognition rate up to 88.1%. Compared with other methods, the proposed models and techniques are believed to have a sufficient potential for a successful application to gait recognition.

Analyzing human interactions with a network of dynamic probabilistic models

In this paper, we propose a novel method for analyzing human interactions based on the walking trajectories of human subjects. Our principal assumption is that an interaction episode is composed of meaningful smaller unit interactions, which we call ‘sub-interactions.’ The whole interaction is represented by an ordered concatenation or a network of sub-interaction models. From the experiments, we could confirm the effectiveness and robustness of the proposed method by analyzing the internal work of an interaction network and comparing the performance with other previous approaches.

Priority-based minimum interference path multicast: routing algorithm in optical virtual private network

While the “Virtual Private Network (VPN) over Internet” is cost-effective and flexible, it suffers from the difficulty of providing adequate transmission capacity for high bandwidth services. Hence a Dense Wavelength Division Multiplexing (DWDM) based “Optical VPN (OVPN)” technology has been regarded as a good alternative for realizing the future VPN services. To improve the transparency and data rate of OVPN, it is critical to consider the problem of Routing and Wavelength Assignment (RWA) for transmission capacity utilization. This paper proposes a Priority-based Minimum Interference Path Multicast Routing (PMIPMR) algorithm, a new routing algorithm which finds alternative routes based on node priorities and Virtual Source (VS) nodes that has both splitting and wavelength conversion, and then chooses a path that does not interfere with potential future multicast session requests when congestions occur in the network. The PMIPMR algorithm reduces blocking rate significantly and increases the wavelength utilization by avoiding congestion in future multicast session requests. We measured the performance of the proposed algorithm in terms of blocking rate and the resource utilization. The simulation results demonstrate that the PMIPMR algorithm is superior to the previous multicast routing algorithms using the Capability-based-Connection algorithm based on Capability-based-Priority and Spawn-from-VS methods.

Human Gait Analysis Using Self Organizing Map

Recently human gait has been considered as a useful biometric supporting high performance human identification systems. Here we are more interested in understanding the gait including the direction rather than the human identity. We propose the use of SOM for interpreting human gait activity using the silhouettes of a pedestrian. The technique and the result may be able to find applications in view-independent direction and gait recognition as well as activity analysis.

Dynamic Character Model Generation for Document Keyword Spotting

This paper proposes a novel method of generating statistical Korean Hangul character models in real time. From a set of grapheme average images we compose any character images, and then convert them to P2DHMMs. The nonlinear, 2D composition of letter models in Hangul is not straightforward and has not been tried for machine-print character recognition. It is obvious that the proposed method of character modeling is more advantageous than whole character or word HMMs in regard to the memory requirement as well as the training difficulty. In the proposed method individual character models are synthesized in real-time using the trained grapheme image templates. The proposed method has been applied to key character/word spotting in document images. In a series of preliminary experiments, we observed the performance of 86% and 84% in single and multiple word spotting respectively without language models. This performance, we believe, is adequate and the proposed method is effective for the real time keyword spotting applications