Voice prompts for use in speech-to-speech translation system

Abstract

Techniques for employing improved prompts in a speech-to-speech translation system are disclosed. By way of example, a technique for use in indicating a dialogue turn in an automated speech-to-speech translation system comprises the following steps / operations. One or more text-based scripts are obtained. The one or more text-based scripts are synthesizable into one or more voice prompts. At least one of the one or more voice prompts is synthesized for playback from at least one of the one or more text-based scripts, the at least one synthesized voice prompt comprising an audible message in a language understandable to a speaker interacting with the speech-to-speech translation system, the audible message indicating a dialogue turn in the automated speech-to-speech translation system.

Description

[0001] This invention was made with Government support under Contract No.: N66001-99-2-8916 awarded by DARPA BABYLON. The Government has certain rights in this invention.FIELD OF THE INVENTION [0002] This present invention generally relates to speech processing techniques and, more particularly, to techniques for employing voice prompts in a speech-to-speech translation system. BACKGROUND OF THE INVENTION [0003] Multilingual speech-to-speech language translation systems have been developed to facilitate communication between people that do not share a common language. One example of such a system is the speech-to-speech translation system developed by Carnegie Mellon University (Pittsburgh, Pa.). [0004] A speech-to-speech translation system allows a user who has been trained with the system (hereinafter “system user”) to communicate with another person who speaks another language (hereinafter “foreign language speaker” or just “foreign speaker”) and is most often not familiar with t...

AI Technical Summary

Benefits of technology

[0015] Accordingly, principles of the invention provide a prompt solution for use in a speech-to-speech translation system that can sufficiently indicate both the switch of dialogue turns and the specific source language for the next turn.

Problems solved by technology

The drawback of this approach is that both the system user and the foreign language speaker need to be trained to be familiar with the meaning of these sound effects.

For a frequent system user, this brings additional inconvenience, as he or she must remember the meaning of sound effects for each language supported by the system.

For a foreign speaker who is not familiar with or has never used this kind of system before, this function is not easily usable for them since the system user cannot explain the function to the foreign speaker because of the language barrier.

The foreign speaker needs to guess the meanings of these sounds, often with great frustration and, consequently, with great dissatisfaction.

However, visual prompts are only helpful in face-to-face speech translation conversations and are useless for other scenarios such as automatic speech translation through call centers.

Additionally, in some situations such as emergency medical care, patients speaking another language may keep their eyes closed due to their medical conditions so that the above-described visual prompts may be completely useless.

Furthermore, these visual indications may still be confusing without verbal explanations.

Images