The Recurrent Sticky Hierarchical Dirichlet Process Hidden Markov Model

TL;DR

This paper introduces the recurrent sticky HDP-HMM, a more flexible Bayesian nonparametric model for spatio-temporal data that improves segmentation performance over previous sticky HDP-HMM variants.

Contribution

It develops the recurrent sticky HDP-HMM, extending previous models by relaxing stationarity assumptions and providing an efficient Gibbs sampling inference method.

Findings

RS-HDP-HMM outperforms previous models in segmentation tasks.

The new model demonstrates improved flexibility and accuracy.

Efficient inference is achieved through a novel Gibbs sampling strategy.

Abstract

The Hierarchical Dirichlet Process Hidden Markov Model (HDP-HMM) is a natural Bayesian nonparametric extension of the classical Hidden Markov Model for learning from (spatio-)temporal data. A sticky HDP-HMM has been proposed to strengthen the self-persistence probability in the HDP-HMM. Then, disentangled sticky HDP-HMM has been proposed to disentangle the strength of the self-persistence prior and transition prior. However, the sticky HDP-HMM assumes that the self-persistence probability is stationary, limiting its expressiveness. Here, we build on previous work on sticky HDP-HMM and disentangled sticky HDP-HMM, developing a more general model: the recurrent sticky HDP-HMM (RS-HDP-HMM). We develop a novel Gibbs sampling strategy for efficient inference in this model. We show that RS-HDP-HMM outperforms disentangled sticky HDP-HMM, sticky HDP-HMM, and HDP-HMM in both synthetic and real data segmentation.