Speech recognition method based on STO array and dynamic time planning algorithm
A dynamic time planning and speech recognition technology, applied in speech recognition, speech analysis, instruments, etc., can solve problems that remain in the theoretical stage, are difficult to apply in practice, and complex implementation, so as to reduce mutual influence, improve reliability and recognition rate, the effect of avoiding the influence
Pending Publication Date: 2022-08-09
HANGZHOU DIANZI UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Regarding the application of STO in the field of speech recognition, some researchers have proposed that STO can be applied to speech recognition by combining STO with neural network storage calculations. However, the disadvantage of this method is that it is too complicated and difficult to implement, and it is only theoretical stage, whether it can be realized is still open to question, so it is difficult to carry out practical application
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 Construction
[0027] The present invention will be further explained below in conjunction with the accompanying drawings;
[0028] figure 1 Schematic diagram of the structure of the spin nano-oscillator designed and used in this embodiment. The diameter of the STO is 600 nm, and it is generally divided into three layers, from top to bottom: the free layer, the non-magnetic layer and the fixed layer. The material is CoFeB, the height is 12 nm, and the magnetic moment density is 5 nm; the material of the non-magnetic layer is Mgo, and the height is smaller than that of the free layer; the material of the fixed layer is CoFeB, and the height and magnetic moment density are the same as those of the free layer. This size can ensure stable oscillation of the STO, and the magnetic moment does not reverse.
[0029] In the micromagnetic simulation software OOMMF, the above STO structure is written into a program in the form of code, and the simulation results are recorded stably oscillates without ...
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 speech recognition method based on an STO array and a dynamic time planning algorithm. The method comprises the following steps of: processing a voice signal by using an STO array; and identifying by using an improved dynamic time planning algorithm. The STO can generate stable microwave signals without an external magnetic field. The method comprises the following steps: firstly, carrying out time domain preprocessing on a signal, then carrying out fast Fourier transform, carrying out calculation to convert a power spectrum into a current signal in a segmented manner, and introducing the current signal into an STO array to obtain an output frequency corresponding to each segment of current as a characteristic vector; and finally, the feature vector is processed by using an improved dynamic time planning algorithm to realize identification. According to the method, an array constructed by a nanoscale STO device is utilized, the characteristics of non-magnetic field overturning, high frequency adjustability and the like of the STO device are combined, an algorithm is improved, the distance is limited, a training library is enlarged, a voice recognition hardware improvement system is realized, and voice recognition with high integration level and high recognition rate is expected to be realized.
Description
technical field [0001] The invention belongs to the technical field of speech signal recognition, relates to a speech signal processing method based on a spin device, and in particular relates to a speech recognition method based on an STO array and a dynamic time planning algorithm. Background technique [0002] The research on speech recognition can be traced back to the middle of the last century. At that time, Bell Labs developed the world's first computer-based speech recognition system. The system can recognize the input simple English pronunciation. Since then, The research on speech recognition all over the world has entered a period of rapid development, and many speech recognition technologies have been invented: in the 1980s, the method of speech recognition using Hidden Markov Model (HMM) was proposed, artificial neural network technology ( ANN) also began to be applied to the field of speech recognition, followed by the emergence of dynamic time planning (DTW) t...
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): G10L15/10G10L25/87G10L21/0208
CPCG10L15/10G10L25/87G10L21/0208
Inventor 崔凯周铁军李海
Owner HANGZHOU DIANZI UNIV