Implementation method and application of a voice wake-up module
A technology of voice wake-up and realization method, applied in voice analysis, voice recognition, mechanical mode conversion and other directions, can solve problems such as changes
- 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 an appropriate database, and obtains it based on MLE (MaximumLikelihoodEstimation, maximum likelihood estimation) and MPE (MinimumPhoneError, minimum phoneme error) discriminative train...
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