Computationally efficient speech extraction algorithms have significant potential economic benefit, by automating an extremely tedious manual process. Previously, algorithms which discriminate between speech and one specific other signal type have been developed, and often fail when the specific non-speech signal is replaced by a different signal type. Moreover, several such signal specific discriminators have been combined in order to tackle the general speech vs. non-speech discrimination problem, with predictable negative results. When the number of discriminating features is large, compression methods such as principal components have been applied to reduce dimension, even though information may be lost in the process. In this paper, graphical tools are applied to determine a set of features which produce excellent speech vs. nonspeech clustering. This cluster structure provides the basis for a general speech vs. non-speech discriminator, which significantly outperforms the TALKATIVE speech extraction algorithm.

This paper presents the Variable Geometry Automated Universal Array (VGAUA) fully automatic physical layout program for gate array LSI which uses a single unique metallization pattern to personalize the array design. This fully automatic layout system minimizes the cost and time required for IC design and can be used over a range of array sizes from 800 to 3500 gates. Each array size has its own unique, optimized, fixed geometry design for all levels except the programmable metallization level.

The philosophy, structure, and design of a data-management system that is applicable to a wide range of organizational environments - from production groups to research and development groups - are introduced and discussed. The system is in operation in two locations and has been used to design numerous LSI/VLSI chips. This low cost system is transportable, extendable, and suitable for use by designers having a diversity of CAD-related skill levels. User reaction to the system and future plans are discussed.