Speech interaction method and device

Abstract

The present invention discloses a speech interaction method. The method comprises: when a first control instruction is received, enabling a speech collection function, outputting a first speech prompting, and starting a speech collection progress icon to allow the speech collection progress icon to move along a setting direction; when the speech collection progress icon is moved along the setting direction, performing collection of speech signals in a current environment; when speech signals are collected before the speech collection progress icon is moved to a limitation position along the setting direction, analyzing the speech signals, and obtaining speech data; matching the speech data and the instruction data in a local instruction bank; and when determining that the speech data and the instruction data in the local instruction bank are successfully matched, outputting a second speech prompting corresponding to the instruction data, and executing a speech instruction corresponding to the speech data. The present invention further discloses a speech interaction device.

Description

Technical field [0001] The invention relates to voice interaction technology, in particular to a voice interaction method and device. Background technique [0002] Voice control technology is a cutting-edge topic in the field of smart device control in the world today. The purpose is to allow the device to accurately perform predetermined actions in accordance with human passwords. At present, the world's major information technology (IT, Information Technology) companies have launched their own speech recognition engines, such as Apple's SIRI, Google's GoogleNow, and Microsoft's Cortana. Domestic IT companies have also launched their own voice recognition services, such as Baidu Voice Assistant and so on. The launch of these voice platforms demonstrated the magical power of voice device control. Devices began to understand people's words and act according to our wishes. In the prior art, the voice control system generally includes the following two methods for acquiring voice ...

AI Technical Summary

Problems solved by technology

However, the defects of the human-machine voice interaction method are: in the far-field interaction, the quality of the voice interaction is greatly affected by the environment. Environmental noise, accent, and volume all directly affect the accuracy of voice recognition, and the application occasions are greatly restricted.

And after the system receives the voice input, there is no voice confirmation link, and the user does not know whether the command recognized by the system is the command sent by the user

Images