Voice endpoint detection model construction method and device, and computer equipment
The invention relates to a voice endpoint detection model construction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring an amplitude spectrum of a training voice signal with noise and a corresponding prior signal-to-noise ratio reference...
Gespeichert in:
Hauptverfasser: | , , , , , |
---|---|
Format: | Patent |
Sprache: | chi ; eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The invention relates to a voice endpoint detection model construction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring an amplitude spectrum of a training voice signal with noise and a corresponding prior signal-to-noise ratio reference value, determining first loss corresponding to a first sub-model according to the amplitude spectrum and the corresponding prior signal-to-noise ratio reference value, adjusting model parameters of the first sub-model according to the first loss to determine a target first sub-model, obtaining a second loss corresponding to the second sub-model according to the frequency characteristic information of the training voice signal with noise, the corresponding voice segment reference value and the non-voice segment reference value, and adjusting the model parameters of the second sub-model based on the second loss to obtain a target second sub-model, and obtaining a target voice endpoint detection model based on fusion |
---|