Metadata-Version: 2.1
Name: voice-analyzer
Version: 1.0.0
Summary: Human voice analyzer
Home-page: https://github.com/artysold/voice-analyzer
Author: Artem Soldatov
Author-email: arty.sold@gmail.com
License: UNKNOWN
Description: # Voice-analyzer
        Human voice analyzer for determining human parameters.
        
        ## Requirements
        - python 3.7 x64 
        - pip (for installation)
        
        ## Installation
        ```
        $ pip install voice-analyzer
        ```
        
        ## Usage
        ### Import
        ```python
        from voice_analyzer import *
        ```
        
        ### Sex
        To determine sex, a ml model trained on the basis of voice frequency is used. 
        The success rate is ~87%.
        ```python
        va_sex = analyzer.Sex()
        ```
        You can use your own model.
        ```python
        va_sex = analyzer.Sex("path\\to\\model.h5")
        ```
        
        #### Prediction
        ```python
        va_sex.predict("path\\to\\file.mp3")
        ```
        The function returns the gender as a string:
        - male (0)
        - female (1)
        
        #### Train
        You can train the model based on your own dataset.
        ```python
        from voice_analyzer import *
        import pandas as pd
        
        va_sex = analyzer.Sex()
        data = [
            {'sex': 0, 'frequency': 180},
            {'sex': 1, 'frequency': 220}
        ]
        df = pd.DataFrame(data)
        
        va_sex.train(df, "path\\to\\save\\model.h5")
        ```
        #### Tests
        You can check the percent accuracy of the model on your dataset.
        ```python
        from voice_analyzer import *
        import pandas as pd
        
        va_sex = analyzer.Sex()
        data = [
            {'sex': 0, 'frequency': 180},
            {'sex': 1, 'frequency': 220}
        ]
        df = pd.DataFrame(data)
        
        va_sex.tests(df)
        ```
        
        
Platform: UNKNOWN
Description-Content-Type: text/markdown
