
Speech--to--Speech Translation: a Massively Parallel Memory-Based Approach describes one of the world's first successful speech--to--speech machine translation systems. This system accepts speaker-independent continuous speech, and produces translations as audio output. Subsequent...