S12.3 JANUS: A SPEECH-TO-SPEECH TRANSLATION SYSTEM USING CONNECTIONIST AND SYMBOLIC PROCESSING STRATEGIES

We present JANUS, a speech-to-speech translation system that utilizes diverse processing strategies including dynamic pro-gramming, stochastic techniques, connectionist learning, and traditional AI knowledge representation approaches. JANUS translates continuously spoken English utterances into Japanese and German speech utterances. The overall system performance on a corpus of conference registration conversations is 87%. Two versions of JANUS are compared: one using an LR parser (JANUS 1) and one using a connectionist parser (JANUS2). Per-formance results are mixed, with JANUS1 deriving benefit from a tighter language model and JANUS2 benefitting from greater flexibility. 1