Layer-by-layer channel selection method for voice recognition of self-organizing microphone
A channel selection and speech recognition technology, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as unhelpful performance, increased network calculation, and no exploration of channel selection, so as to improve recognition performance and reduce computational complexity. Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0128] DETAILED DESCRIPTION
[0129] This embodiment uses three data sets: librispeech corpus, libri-adhoc-simu dataset in the self-organized microphone array environment based on librispeech simulation, and 40 distributed microphones play back librispeech's libri-adHoc40 in real environments. Each node of the self-organized microphone array of libri-adhoc-simu and libri-adhoc40 is a single microphone, a channel represents a node. Librispeech contains 2484 speakers more than 1,000 hours of English speeches. In the example, 960 hours of data were selected to train the single channel ASR system and selected 10 hours of data for verification.
[0130] For simulation data, libri-adhoc-simu uses librispeech data for 100 hours "Train-100" subset as training data. Use the "dev-clean" subset as the verification data, a total of 10 hours of data. The "Test-Clean" subset is used as two separate test sets, which contain 5 hours of test data, respectively. The length and width of the simulate...
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