Robust Classification using Hidden Markov Models and Mixtures of Normalizing Flows

TL;DR

This paper introduces NMM-HMM, a generative model combining HMMs and neural network-based distributions, which enhances robustness in sequential data classification tasks like speech recognition by integrating normalizing flows.

Contribution

The paper proposes the NMM-HMM model that integrates hidden Markov models with neural network-based probability distributions using normalizing flows, trained via EM and backpropagation.

Findings

NMM-HMM improves robustness in speech recognition.

The model effectively combines HMMs with neural network distributions.

Experimental results show enhanced noise resilience.

Abstract

We test the robustness of a maximum-likelihood (ML) based classifier where sequential data as observation is corrupted by noise. The hypothesis is that a generative model, that combines the state transitions of a hidden Markov model (HMM) and the neural network based probability distributions for the hidden states of the HMM, can provide a robust classification performance. The combined model is called normalizing-flow mixture model based HMM (NMM-HMM). It can be trained using a combination of expectation-maximization (EM) and backpropagation. We verify the improved robustness of NMM-HMM classifiers in an application to speech recognition.