Initialization-robust Bayesian multipitch analyzer based on psychoacoustical and musical criteria

We present a new Bayesian multipitch analyzer that dispenses with a precise optimization of parameter initialization or hyperparameters. Our method uses a new family of prior distribution, characteristic prior; it efficiently restricts the existence region of the latent variables, that is, the product of a conjugate prior and a characteristic function. The update formulas become a simple form that is actually suitable for Gibbs sampling. We construct characteristic priors of harmonic structures based on psychoacoustical and musical knowledge and apply them to nonnegative harmonic factorization. Experimental results improve 5.2 points in F-measure under a tough condition, random initialization with no hyperparameter optimization.