Voiceprint recognition method, device and apparatus for original voice, and storage medium
A voiceprint recognition, the original technology, applied in voice analysis, instruments, etc., can solve the problems of high information loss and high system complexity
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0069] The embodiment of the present invention provides a method, device, device and storage medium for recognizing voiceprint of original speech. Through a new loss function, the noise information in the channel for recognizing voiceprint characteristic information in the original speech data is eliminated, and the information is reduced. Loss, through the use of a preset convolution filter bank as the front-end pre-processing structure of the original voice data to obtain voiceprint feature data, and a preset deep neural network to pool the voiceprint feature data, through the cosine similarity matrix Loss function and minimum mean square error matrix The loss function processes the voiceprint feature vector to obtain target voiceprint data in the form of a similarity matrix or embedding vector. The input is the original speech data of the speaker and the output is the similarity matrix or embedding vector. The structural form of the target voiceprint data in the form of the s...
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.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap