Construction method and device of voice recognition method
A technology of speech recognition and corpus, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as failure
Active Publication Date: 2019-11-19
AISPEECH CO LTD
View PDF8 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] In the process of implementing this application, the inventor found that both the online speech recognition of a certain company and the online speech transcription of another company require a networked environment, and will fail when the network signal is not good or there is no network
Another company's offline command word recognition, although it can support disconnection recognition, only supports a small number of command word recognition
It also fails if the speaker's speech content is outside the command word
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment approach
[0061] As an implementation manner, the above-mentioned electronic equipment is applied to a construction device of a speech recognition system, including:
[0062] at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions are executed by the at least one processor so that the at least one processor can:
[0063] Generate navigation corpus based on popular navigation vocabulary and navigation statements;
[0064] Combining the navigation corpus with the corresponding local vocabulary based on a preset toolkit to generate a navigation language model;
[0065] performing preset processing on the navigation language model and the preset small domain model to form a local language model of a predetermined size;
[0066] A speech recognition system is generated using the local language model, pronunciation dictionary and acoustic model.
[0067] El...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a construction method and device of a voice recognition system. The construction method of the voice recognition system comprises the following steps: generating navigation linguistic data based on high-popular navigation vocabulary and a navigation statement; combining the navigation linguistic data with the corresponding local vocabulary based on a preset toolkit to generate a navigation language model; preprocessing the navigation language model and the preset small domain model to form the local language model with predetermined size; and generating a local voice recognition system by combining the local language model with a pronunciation dictionary and an acoustic model. The resources provided by the method and device can be installed on a user vehicle-mountedproduct to use, and the following demands are basically satisfied: moderation of the resource guaranteeing, the low power consumption and low internal memory, and moderate recognition performance guaranteeing.
Description
technical field [0001] The invention belongs to the technical field of speech recognition, and in particular relates to a construction method and device of a speech recognition system. Background technique [0002] Among related technologies, speech recognition technology is more and more widely used in daily life, which brings great convenience to people's life. Among them, one company's online speech recognition, another company's online speech transcription and offline command word recognition are several common technologies at present. [0003] Among them, the online speech recognition of a certain company is a product that calls the speech recognition SDK that needs to be connected to the Internet; the online speech transcription of another company is also a product that calls the speech recognition SDK that needs to be connected to the Internet; the offline command word recognition of another company can case, the recognition of command words. [0004] The inventor f...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G10L15/06
CPCG10L15/06G10L15/063G10L2015/0631
Inventor 胡雪成
Owner AISPEECH CO LTD