Skip to content

🥈 2nd Place - Eestech Challange LC Trieste - Hackathon Project

License

Notifications You must be signed in to change notification settings

Fedrosauro/KitarKit

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

KitarKitLogo2


Project Group:

  • Emanuele Goat
  • Federico Pellizzaro
  • Giovanni Zanin

Task:

AI-based tool for supporting artists' creativity

Example:

  1. Generation of book or music album covers
  2. Book illustrations
  3. Songwriting
    ...

Solution proposed:

For the specific case of guitarists, a model has been trained to recognize which effects have been applied to a guitar sound. The best accuracy found on the test set was ~75%.

Due to the dataset limitations, the model in this repo can currently recognize just one effect applied to a guitar sound. However, the CNN has been studied and developed so that a model can be trained with a dataset where multiple sound effects are applied to a guitar sound.

Additionally, a small web app has been developed to showcase the model's response when given random guitar sounds sourced from the internet.

In order to have the web app running you need conda and then:

  1. conda env create -f environment.yml
  2. conda activate kitar-kit-env
  3. streamlit run .\KitaraKitWebApp.py

EC_logo__white_bg__1200_px