Audio conversation device, method, and robot device

Abstract

In conventional voice dialogue devices, there are cases where it is difficult to have a natural dialogue with a user. Carry out voice recognition on the user's voice, and control the dialogue with the user according to the predetermined scene according to the result of the voice recognition, and, if necessary, generate a response sentence according to the content of the user's voice, so as to audio synthesize a sentence for reproducing the scene Or generated response sentences.

Description

Technical field [0001] The present invention relates to a voice dialogue system and method and robot equipment, for example, a robot suitable for entertainment. Background technique [0002] The method of dialogue with people through voice by the voice dialogue system can be divided into two categories according to the content. They are "scenario dialogue" and "scenario dialogue". [0003] Among the two types of methods, the "scenarioless dialogue" method is a dialogue method called "non-artificial intelligence", which is realized by a simple answer sentence generation algorithm represented by E1iza (see Non-Patent Document 1 ). [0004] In the "dialogue without scene" method, as shown in Figure 36, processing is performed by repeating an iterative process (step SP92). If the user utters a few words, the voice dialog system performs voice recognition on these words (step SP92). SP90), and generate a reply sentence based on the recognition result, and utter the reply sentence by v...

AI Technical Summary

Problems solved by technology

[0010] However, this dialogue method also has some problems

First of all, since the speech dialogue system can only speak according to the pre-designed scene assuming the user's reply content, the speech dialogue system cannot respond when the user speaks something unexpected

[0011] For example, for a question that can be answered with "yes / no", if the user answers both, but he has never thought of such a thing or something similar, the voice dialogue system cannot make any response, even if he does response, and only a highly inappropriate response to the user's reply

Furthermore, in such cases, the likelihood that the plot will become unnatural later on increases

[0012] Second, it is difficult to set the degree of occurrence ratio of the two rounds where the voice dialogue system speaks unilaterally and the voice dialogue system asks the user a question and then responds based on the user's answer to the question

[0013] In fact, in the above-mentioned voice dialogue system, if the previous round is too frequent, it will create the impression that the voice dialogue system speaks to the user unilaterally, and the user does not feel that "the dialogue is in progress"

Conversely, if the latter round is too frequent, it creates the feeling that the user is responding to a survey or query of the user; in this case, the user also does not feel that a "conversation is ongoing"

Images