Application of fuzzy logic in order to plan and create software for converting sound to text

Hiwa Hasanzadeh, Jafar Majidpour

Abstract


The challenge of software in the field of converting sound to text is sometimes the problems of recognizing and converting sound and words. Some words have different meanings and may have problems with the translation of the machine. In these situations, the functions of each one are distinguished by the use of fuzzy logic and division of the working part of the work. In this regard, the words must first be classified according to the specialized discipline. For example, for basic sciences and engineering and medicine, the meaning of a specific word can be exactly defined in each field, and the type of string is specified when entering data. The better the split, the better. In the second place, the principles of writing should be maintained. In this regard, the points and commands and the alignment of the sentence are all items that should be considered in the software. The next item is the volume and speed of word processing and the lack of restrictions in the number of languages used. But the time problem is very important as one of a variety of software developed by the algorithm is such that you can convert a 1-minute audio file to the text in 5 seconds. In this regard, the use of existing codes and the optimization and remedy of the problems is a suitable method.In this research, some issues including security and maintenance of software proprietary rights, and the speed of converting sound to text and reducing noise and optimal performance of software to convert audio to text based on software method and the principles of fuzzy logic and neural network were investigated. Results showed improvement algorithm Analytical and results outcomes.

Keywords


Algorithm Analytical; Fuzzy Logic; Neural Network; Speed of Converting Sound to Text; The Challenge of Software


Refbacks

  • There are currently no refbacks.


Bulletin of EEI Stats