Method of checking voice detection result and apparatus thereof

Abstract

The invention provides a method of checking a voice detection result and an apparatus thereof. The method comprises the following steps of if a current frame detection result is that a voice activity exists, calculating a statistical value of a detection result in a detection result buffer area; and comparing the statistical value to a preset threshold, acquiring a determination result. The invention provides a method of checking the voice detection result and the apparatus thereof. A characteristic that a voice generally possesses coherence is considered. After the current frame detection result is calculated, detection results of previous frames are combined to acquire the statistical value. The statistical value is compared to the preset threshold, whether the current frame detection result is misjudged is determined and voice detection accuracy is increased.

Description

technical field [0001] The invention relates to the field of voice recognition, in particular to a method and device for checking voice detection results. Background technique [0002] In recent years, with the vigorous development of Internet technology and intelligent hardware, voice intelligent interactive technologies such as voice recognition, voiceprint recognition, and sound source detection have begun to move from laboratories to users. Because speech recognition technology is the core technology of human-computer interaction system based on speech. At present, the recognition rate has reached the usable accuracy rate under limited conditions. The so-called limited adjustment usually means that the distance between the user and the microphone is relatively close, and the noise interference is small. However, the condition that voice commands must be issued at close range limits the convenience of voice interaction. [0003] In the case of long-distance speaking, s...

AI Technical Summary

Problems solved by technology

[0006] However, this method uses the calculated cepstrum distance to compare with the preset threshold. Although the thresholds at different distances are preset, due to the complexity of the actual environment, the preset threshold often cannot be applied in specific scenarios. lead to a decline in the accuracy of speech recognition

At the same time, this solution does not have a process of checking the voice detection results, which is prone to false detections

Images