Recognising people from their gait is a challenging problem in biometric research. In this paper, we address the problem of gait identification based on a novel approach of sub-vector quantisation (SVQ) technique. A silhouette-based algorithm is utilised to capture the spatial-temporal information of the gait. A sequence of temporally ordered outer contour widths of binarised silhouettes of a w...

Although the continuous hidden Markov model (CHMM) technique seems to be the most flexible and complete tool for speech modelling, it is not always used for the implementation of speech recognition systems because of several problems related to training and computational complexity. Thus, other simpler types of HMMs, such as discrete (DHMM) or semicontinuous (SCHMM) models, are commonly utilise...

-Classification, a data mining task is an effective method to classify the data in the process of Knowledge Data Discovery. Classification method algorithms are widely used in medical field to classify the medical data for diagnosis. Feature Selection increases the accuracy of the Classifier because it eliminates irrelevant attributes. This paper analyzes the performance of neural network class...

In this paper, we present new results on Temporal Decomposition (TD) applied to the Line Spectral Frequencies (LSFs) derived for wideband speech. The paper shows that by incorporating a dynamic programming search algorithm into TD, near transparent quantisation of wideband LSFs can be obtained at approximately 1 kbps. We also show that TD performs significantly better than Split Vector Quantisa...

Image compression is concerned with reducing the amount of data needed to represent an image. Efficient image representation is achieved by exploiting the statistical and psychovisual redundancies of an image. This reported focusses on the main principles of information theory, which provides a framework for efficient signal coding from a statistical perspective. Two of the fundamental theories...

We present a new approach to multiresolution vector quantisation. Its main advantage is exploitation of long-range correlations in the image by keeping vector size constant, independent of the image scale. We also developed a variable block-rate version of the algorithm, which allows better utilisation of the available bit budget by re ning only those areas of the image which are not e ciently ...

A low hit rate coding method using the discrete wavelet transform (DWT), and classified vector quantisation (CVQ) is presented. An image is decomposed by DWT; then, the transformed image is encoded by CVQ. The proposed method allows the decoder to self-generate the class information for CVQ rather than requiring that it be received from the encoder. This results in improved compression performa...

 A constrained SOM based on an equal-distortion principle is proposed for producing globally optimal, or near-optimal, vector quantisation. The principle is applied indirectly to control the width of the neighbourhood of the SOM. Little extra computation costs are introduced but improved performance, both in lower distortion and in stable and fast convergence, is achieved.