For example, on Youtube you can find several program videos that shows which chords are being played at the music. Of course, a lot of people thought the same thing. Then I thought, instead of letting the machine do everything, we could use the computer just to help music students to understand better musical compositions. Since neural networks have “hidden layers”, it becomes harder to understand why the machine “choose” to create the music using a specific structure. Once I was reading about music generation using neural networks and remembered another article about Explainable Artificial Intelligence. Main Objective: Computer Aided Music Analysis You can find all code on my Kaggle Kernel link. In this article I’ll describe in high level some results which I achieved. This way I decided to create a new kernel there to learn more about MIDI files and play with the Word2Vec over harmonic sequences. There are few databases and kernels and most of them are focused on analyzing music recommendation and preferences. I was taking a look there to see if I could find anything related with music or MIDI files but I didn’t find anything. Users can create and edit Kernels online to share knowledge about some topic like machine learning algorithms or Python data structures. One of those feature is an online editor of scripts and notebooks which are called of “Kernel” by them. Kaggle is a community site for data scientists and machine learners which offers several features to support their work and research. MIDI Music Data Extraction using Music21 and Word2Vec on Kaggle
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |