Special Issue on VLSI Computing for Real-World Intelligent Systems
6-6-05 Aoba, Aramaki, Aoba, Sendai, Miyagi 980-8579, Japan
Graduate School of Information Science, Tohoku University
Recently, intelligent systems are desired to support human in real world such as advanced safe vehicles, home service robots, wearable computing devices, and intelligent home security systems. Such intelligent systems require extremely high computational power that exceeds that of state-of-the-art microprocessors. They also require
- Low power consumption
- Low latency from input to output
Special-purpose processors called “system LSIs” play an essential role in meeting these requirements. This special issue focuses on the latest advances in system LSIs for real-world intelligent systems. One of their most important tasks is sensing environmental information such as visual information. Image and angle sensors, for example, are implemented in system LSIs. Image processing is the most time-consuming in real-world intelligent systems due to the extremely large amount of data. To overcome this problem, novel parallel architectures are presented. Electrical wires between processing modules must be minimized to make intelligent systems compact. High-speed serial data transfer is one most effective way to minimize the electrical wires. An architecture that handles processing order based on task priorities is a key to low latency. Processing of human interfaces such as face detection and speech recognition are also important factors in making intelligent systems user-friendly. I thank the authors of the articles in this issue for their effort and contributions, and the members of the Editorial Board for their cooperation.
Copyright© 2005 by Fuji Technology Press Ltd. and Japan Society of Mechanical Engineers. All right reserved.