Music infilled by different models

Examples of the infilling result of our model and the two baselines when the past and the future contexts are four bars apart. While our model can indeed infill the missing segment with four bars long as expected, both ILM and FELIX tend to generate much shorter segments.

Note: If you are browsing this page using an iPhone, please remember to turn the silent mode off through the switch on the side of your iPhone.

Song 01

Song 02

Song 03

Song 04

Music infilled by our model to have a different number of bars

We can decide how many bars we want our model to infill in the missing segment. Here, we provide examples of our models to generate 1 bar to 4 bars to connect the contexts.

Song 01

Song 02

Song 03

Song 04