In silico vox: Towards speech recognition in silicon

Abstract

This article consists of a collection of slides from the author's conference presentation on Carnegie Mellon University's In Silico Vox speech recognition in silicon Project. Some of the specific topics discussed include: current speech recognition technologies; the special features and voice commuication capabilities of the In Silico Project; graphics engines; next generation speech recognition applications; how speech recognition works; acoustical capabilities; layered speech models; software design and implementation; pipelining facilities; an FPGA demonstration and experimental results.