18557213. Method and System for Constructing Speech Recognition Model and Speech Processing simplified abstract (Huawei Technologies Co., Ltd.)
Contents
Method and System for Constructing Speech Recognition Model and Speech Processing
Organization Name
Inventor(s)
Method and System for Constructing Speech Recognition Model and Speech Processing - A simplified explanation of the abstract
This abstract first appeared for US patent application 18557213 titled 'Method and System for Constructing Speech Recognition Model and Speech Processing
- Simplified Explanation:
The patent application describes a method for constructing a speech recognition model by training a language model based on a target keyword and its associated synonym group, generating a decoding graph based on the language model, and determining the speech recognition model based on the decoding graph.
- Key Features and Innovation:
- Obtaining a target keyword and determining a synonym group associated with it - Training a language model based on the keyword and synonym group - Generating a decoding graph with multiple paths based on syntax constraints - Determining the speech recognition model based on the decoding graph
- Potential Applications:
- Speech recognition technology - Natural language processing - Voice-controlled devices - Virtual assistants
- Problems Solved:
- Improving speech recognition accuracy - Enhancing language model training - Increasing the efficiency of decoding processes
- Benefits:
- Enhanced speech recognition performance - Improved user experience with voice-controlled systems - More accurate transcription of spoken language
- Commercial Applications:
- This technology can be applied in various industries such as telecommunications, customer service, healthcare, and automotive for developing advanced speech recognition systems.
- Questions about Speech Recognition Model Construction:
1. How does the method in the patent application differ from traditional speech recognition model construction techniques? 2. What are the potential limitations of using synonym groups in training language models for speech recognition?
Original Abstract Submitted
A method for constructing a speech recognition model includes obtaining a target keyword; determining a synonym group semantically associated with the target keyword; training a language model based on the target keyword and the synonym group; to obtain a target language model; generating a first decoding graph based on the target language model, where the first decoding graph indicates a plurality of decoding paths that satisfy a syntax constraint rule that is based on the target keyword and the synonym group; and determining the speech recognition model based on the first decoding graph.
