Unlock instant, AI-driven research and patent intelligence for your innovation.
Scene recognition method and device based on channel characteristics
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A channel feature and scene recognition technology, applied in voice analysis, instruments, etc., can solve problems affecting the classification of sound scenes
Pending Publication Date: 2021-03-12
深圳市云网万店科技有限公司
View PDF11 Cites 2 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0006] In order to solve the problems of the prior art, the embodiment of the present invention provides a scene recognition method, device, computer equipment and storage medium based on channel characteristics, so as to overcome the problem existing in the prior art that the mixed human voice in the audio to be recognized will affect the sound. Scene classification and other issues
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 more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0058] Specifically, the above solution can be realized through the following steps:
[0059] Step 1. Record training audio data in different scenarios
[0060] Specifically, the training audio data is recorded in different channels, and the training audio data includes audio data recorded by different users in different channels. For example, including I different channel scenarios, K different users record J pieces of different audio data in each channel scenario. As a better example, the length of each piece of audio data is about 10s, the volume of the recorded sound remains unchanged, and the audio data is numbered, for example:
[0061] A 11 Indicates the first audio of user A in channel scene 1,
[0062] K ij Represents the jth audio of user K in channel scene i, etc.
[0063] Step 2. Use all the collected audio to train the recognition model
[0064] Specifically, use all the collected training audio data to train a stable recognition model, such as a high-order ...
Embodiment 2
[0099] figure 1 is a flow chart of a method for scene recognition based on channel features according to an exemplary embodiment, refer to figure 1 As shown, the method includes the following steps:
[0100] S1: Acquire audio to be identified, perform frame division processing on the audio to be identified, and acquire audio frames.
[0101] Specifically, in order to facilitate subsequent extraction of feature data of the audio to be recognized, after the audio data to be recognized is obtained, the audio to be recognized needs to be processed by frame division. In the embodiment of the present invention, the frame division processing method includes but is not limited to adopting methods such as matlab.
[0102] S2: Extracting audio features and channel features of each audio frame, and respectively constructing an audio feature vector and a channel feature vector of each audio frame according to the audio features and the channel features.
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 scene recognition method and device based on channel characteristics, computer equipment and a storage medium; and the method comprises the steps: obtaining a to-be-recognized audio, carrying out framing of the to-be-recognized audio, obtaining audio frames, and extracting the audio characteristics and channel characteristics of each audio frame; respectively constructingan audio feature vector and a channel characteristic vector of each audio frame, inputting each audio frame into a pre-trained identification model, obtaining a super-vector corresponding to each audio frame, and performing iterative computation by using a preset method according to the super-vectors, the audio characteristic vectors and the channel characteristic vectors; and obtaining a channelspace matrix corresponding to the to-be-recognized audio, calculating the similarity between the to-be-recognized audio and the to-be-matched audio according to the channel space matrix and the channel space matrix of the to-be-matched audio, and determining a scene corresponding to the to-be-matched audio of which the similarity meets a preset condition as the scene of the to-be-recognized audio. Therefore, the interference of speaker information in a scene recognition process is effectively avoided.
Description
technical field [0001] The present invention relates to the technical field of speech processing, in particular to a scene recognition method, device, computer equipment and storage medium based on channel characteristics. Background technique [0002] Sound scene classification is a technology that processes the sound features of different scene environments to make scene judgments. [0003] Nowadays, the Internet and computer technology are undergoing earth-shaking changes every day. Smart terminals are the entrance for people to interact with the Internet or computers. Voice recognition is undoubtedly the most convenient way for human-computer interaction. The smart terminal may be in various scenarios, and the noise in the environment will greatly interfere with the human voice. Classifying the environmental information in the audio in advance and performing targeted front-end signalprocessing can greatly improve the accuracy of later recognition. [0004] In the exis...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More