Layer-by-layer channel selection method for voice recognition of self-organizing microphone
一种通道选择、语音识别的技术,应用在语音识别、语音分析、仪器等方向,能够解决性能没有帮助、网络计算量增大、没有探究通道选择等问题,达到识别性能提升、降低计算复杂度的效果
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Examples
specific Embodiment
[0129] This embodiment uses three data sets: the Librispeech corpus, the Libri-adhoc-simu data set based on the Librispeech simulation obtained under the self-organizing microphone array environment, and the Libri-adhoc40 in which 40 distributed microphones play back Librispeech in a real environment. Each node of the self-organizing microphone array of Libri-adhoc-simu and Libri-adhoc40 is a single microphone, and one channel represents one node. Librispeech contains more than 1000 hours of English speeches by 2484 speakers. In the embodiment, 960 hours of data are selected to train the single-channel ASR system, and 10 hours of data are selected for verification.
[0130] For the simulation data, Libri-adhoc-simu uses the 100-hour "train-100" subset of the Librispeech data as training data. Use the “dev-clean” subset as validation data, containing a total of 10 hours of data. Treat the "test-clean" subset as two separate test sets, each containing 5 hours of test data. Th...
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