Ai Dreams Forum

Member's Experiments & Projects => General Project Discussion => Topic started by: chung on October 15, 2019, 07:11:04 am

Title: wavtomid chung polyphonic audio wav to midifile converter using neural network
Post by: chung on October 15, 2019, 07:11:04 am

wavtomid_chung : i noticed that the spectrum harmonics correction of my wav to mid converter has the same structure as a single hidden layer neural network , so i tried to generalize it with a randomize optimization algorythm , seems to work better (or is it just an illusion ?)   ::)

=>  http://chungswebsite.blogspot.com/2019/06/wavtomidchung-high-quality-polyphonic.html (http://chungswebsite.blogspot.com/2019/06/wavtomidchung-high-quality-polyphonic.html)
 

Title: Re: wavtomid chung polyphonic audio wav to midifile converter using neural network
Post by: Art on October 17, 2019, 05:50:55 pm

Pretty cool Chung.

The converter did a credible job of forming its take of Glenn Miller's hit, Moonlight Serenade.

Nice work! O0

Title: Re: wavtomid chung polyphonic audio wav to midifile converter using neural network
Post by: goaty on October 17, 2019, 06:08:59 pm

wav2mid is quite complex,  nice project.

Title: Re: wavtomid chung polyphonic audio wav to midifile converter using neural network
Post by: chung on November 13, 2019, 04:12:11 pm

(13/11/2019) new spectrum filter => no need of neural network but may use also  (brings a little gain that can make the difference)  ::)