Volume 20, 2001, No. 3

Automatic Eye Feature Extraction in Human Face Images

Liu RUJIE, Yuan BAOZONG   Abstract This paper presents a fuzzy-based method to extract the eye features in a head-shoulder image with plain background. This method is comprised of two stages, namely the face region estimation and the eye features extraction. In the first stage, a region growing method is adopted to estimate the face region. In the second stage, the coarse eye area is firstly determined based on the location of the nasion, then the deformable template algorithm is completed in two steps to extract the features of the eyes. Experimental results show the efficiency and robustness of this method.

 

A Hybrid Architecture for Multimedia Processors

Bertil SCHMIDT, Heiko SCHRÖDER   Abstract Novel algorithmic features of multimedia applications and System on Chip (SoC) design using state-of-the-art CMOS technology are driving forces behind new multimedia processors. In this paper we propose an architecture that - based on this approaching technology - provides high performance and flexibility. It is a hybrid design consisting of instruction systolic arrays (ISAs) to be used as a special-purpose accelerator and RISC cores to be used as the basis of a general-purpose processor. It is a hierarchical and scalable architecture, which facilitates the hardware-/software codesign of multimedia processing circuits and systems. While some control-intensive functions can be implemented using the general-purpose CPU, other computation-intensive functions can rely on the accelerator.

 

Using Semantical Information to Enhance the Parallel Sparse Performance

Gerardo BANDERA, Emilio L. ZAPATA   Abstract This work presents a novel strategy for the parallelization of applications containing sparse matrix references using the data-parallel paradigm. Our approach is a first step to converge to the automatic parallelization by reducing the number of directives on code. We have used the semantical relationship of vectors composing a high-level data structure to enhance the performance of the parallel code, applying a sparse privatization and a multi-loop analysis. We also study the building/updating of a sparse matrix at run-time, solving the problem of using pointers and some levels of indirections on the left hand side. A detailed analysis about several temporary buffers useful for sparse communications is described in this paper. The evaluation of our strategy has been performed on a Cray T3E with sparse matrix transposition algorithm as a case of study.

 

Towards a Communication-Aware Task Scheduling Strategy for Heterogeneous Systems

Juan Manuel ORDUNA, Federico SILLA, José DUATO   Abstract Many research activities have focused on the problem of task scheduling in heterogeneous systems from the computational point of view. However, a scheduling strategy should also take into account the communication requirements of the applications and the communication bandwidth offered by the network. Towards this end, in this paper we first propose a model of communication cost between network nodes. This model can be used to properly characterize the existing network resources. Second, we propose a criterion to measure the suitability of each allocation of network resources to each parallel application, according to the communication requirements. Third, we propose a scheduling technique based exclusively on this criterion that provides a near-optimal mapping of processes to processors according to the communication requirements. Evaluation results show that the use of this scheduling technique fully exploits the available network bandwidth, greatly improving network performance. Therefore, the proposed scheduling technique can be used in the design of communication-aware scheduling strategies for those situations where the communication requirements are the system performance bottleneck.

 

A Case Based Reasoning System Capturing Fire Modelling Expertise

Stephen TAYLOR, Brian KNIGHT, Miltos PETRIDIS, John EWER, Edwin GALEA   Abstract This paper describes the architecture of the case based reasoning (CBR) component of Smartfire, a fire field modelling tool for use by members of the Fire Safety Engineering community who are not expert in modelling techniques. The CBR system captures the qualitative reasoning of an experienced modeller in the assessment of room geometries so as to set up the important initial parameters of the problem. The system relies on two important reasoning principles obtained from the expert: 1) there is a natural hierarchical retrieval mechanism which may be employed; and 2) much of the reasoning on a qualitative level is linear in nature, although the computational solution of the problem is non-linear. The paper describes the qualitative representation of geometric room information on which the system is based, and the principles on which the CBR system operates.

 

Volume 20, 2001, No. 3