Two Minute Papers has a video exploring the academic paper “The challenge of realistic music generation: modelling raw audio at scale” which details how an AI learned piano from the greatest masters in history.
In keeping with today’s musical theme, Two Minute Papers covers the paper “A Neural Parametric Singing Synthesizer” which talks about creating believable singing given a MIDI file as input data.
In this episode of the AI show Erika follows up on her previous episode by showing the actual code behind training and using the music generation model.
This includes the code for both creating the features from the midi file and returning a midi file from the features. She also shows how Keras is used to generate the actual model.
Learn more:
In this episode of the AI show Erika explains how to create deep learning models with music as the input. She begins by describing the problem of generating music by specifically describing how she generated the appropriate features from a midi file. She then describes the deep learning model she used in order to generate music.
Learn more:
Comments
Deep Fakes and the Technology Behind It
The Code Behind Deep Learning for Music Generation
Explaining Quantum Computing with a Deck of Cards
Android nougat begins rolling out to nexus
Android nougat begins rolling out to nexus