Implementation method and application of voice awakening module
A technology of voice wake-up and implementation method, applied in voice analysis, voice recognition, mechanical mode conversion and other directions, can solve problems such as changes, achieve the effect of better voice interaction, convenient code porting, and good promotion and application value
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0039] Such as figure 1As shown, the realization of the voice wake-up module of the present invention is realized by 3 steps of voice input 1, voice wake-up algorithm 2 and wake-up execution.
[0040] Voice wake-up algorithm 2 is mainly implemented by acoustic feature extraction 4, wake-up word detection 5, wake-up word confirmation 6, construction of wake-up word detection network 7, training acoustic model 8 and construction of wake-up word confirmation network 9. The specific implementation process is:
[0041] (1) Training the acoustic model 8: The training of the acoustic model is divided into two parts, which are the phoneme acoustic model and the waste model (ie Garbage model). The phoneme acoustic model adopts the traditional acoustic model training method in speech recognition, selects a suitable database, and uses the discriminative training criteria based on MLE (Maximum Likelihood Estimation) and MPE (Minimum Phone Error, minimum phoneme error) get. The Garbage m...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com