GAN Improvement

Here we present the improvement between without and with GAN in the proposed model. Without GAN, the model will suffer from "over-smoothing" problem, where the model generates uniform and smoothed harmonic distribution and noise magnitudes over the whole note. See below for an intuitive evaluation.

Ground-truth Harmonic Distribution

Predicted Harmonic Distribution without GAN

With the introduction of adversarial training, the proposed model overcomes the over-smoothing problem potentially caused by one-to-many mapping.

Predicted Harmonic Distribution with GAN

One can also hear the over smoothing problem from predicted samples:

Ground-truth

Without GAN

With GAN

Similar effects can also be seen in noise magnitudes:

Ground-truth Noise Magnitudes

Predicted Noise Magnitudes without GAN

Predicted Noise Magnitudes with GAN