This work aimed to use Random Forest machine learning algorithm for human recognition. Results shown below.
####Brief files description:
- graphs.py - Used for beautiful graphs drawing.
- main.py - Provides experiments.
- pretty_print.py - Prints result pretty.
- recognition.py - Algorithm preparing methods.
- results-*.txt - results of speaker identity and gender recognition.
- speaker-info.txt - Speaker's base information.
###Used dataset could be downloaded from:
http://homepages.inf.ed.ac.uk/jyamagis/page3/page58/page58.html