Learning Musical Pitch Structures with Hierarchical Hidden Markov Models
نویسندگان
چکیده
In this paper we attempt to demonstrate the strengths of Hierarchical Hidden Markov Models (HHMMs) in the representation and modelling of musical structures. We show how relatively simple HHMMs, containing a minimum of expert knowledge, use their advantage of having multiple layers to perform well on tasks where flat Hidden Markov Models (HMMs) struggle. The examples in this paper show a HHMM’s performance at extracting higherlevel musical properties through the construction of simple pitch sequences, correctly representing the data set on which it was trained.
منابع مشابه
A Hierarchical Bayesian Model of Chords, Pitches, and Spectrograms for Multipitch Analysis
This paper presents a statistical multipitch analyzer that can simultaneously estimate pitches and chords (typical pitch combinations) from music audio signals in an unsupervised manner. A popular approach to multipitch analysis is to perform nonnegative matrix factorization (NMF) for estimating the temporal activations of semitone-level pitches and then execute thresholding for making a pianor...
متن کاملMelodic Analysis using Hierarchical Hidden Markov Models
This dissertation attempts to show that hierarchical hidden Markov models (HHMMs) can be used effectively to model mid-level musical structures. Using data taken from chorales composed by Johann Sebastian Bach, a number of models were designed and trained to reflect elements of phrase-level and bar-level structure. These models were evaluated based on their performance in two different tasks; t...
متن کاملHierarchical Sequential Memory for Music: A Cognitive Model
We propose a new machine-learning framework called the Hierarchical Sequential Memory for Music, or HSMM. The HSMM is an adaptation of the Hierarchical Temporal Memory (HTM) framework, designed to make it better suited to musical applications. The HSMM is an online learner, capable of recognition, generation, continuation, and completion of musical structures.
متن کاملLearning Hierarchical Hidden Markov Models for Video Structure Discovery
Structure elements in a time sequence are repetitive segments that bear consistent deterministic or stochastic characteristics. While most existing work in detecting structures follow a supervised paradigm, we propose a fully unsupervised statistical solution in this paper. We present a unified approach to structure discovery from long video sequences as simultaneously finding the statistical d...
متن کاملScale- and Rhythm-Aware Musical Note Estimation for Vocal F0 Trajectories Based on a Semi-Tatum-Synchronous Hierarchical Hidden Semi-Markov Model
This paper presents a statistical method that estimates a sequence of musical notes from a vocal F0 trajectory. Since the onset times and F0s of sung notes are considerably deviated from the discrete tatums and pitches indicated in a musical score, a score model is crucial for improving timefrequency quantization of the F0s. We thus propose a hierarchical hidden semi-Markov model (HHSMM) that c...
متن کامل