A Spoken Dialogue System for Noisy Environment
説明
One of the important challenges for achieving a spoken dialogue system in noisy environments is to make the system’s speech audible for the user. Although there have been many studies on speech recognition in noisy environments, very few attempts to improve the audibility of the system’s speech. In this paper, we develop a spoken dialogue system that has three functions: real-time volume adjustment, utterance delay, and re-utterance. Experimental results have shown that these three functions improve the audibility of the system’s utterances.