A Latent-Variable Lattice Model

TL;DR

This paper introduces a non-Markov lattice model for computer vision tasks that offers faster learning from small datasets by leveraging vector quantization, addressing the intractability of traditional MRFs.

Contribution

The paper proposes a novel non-Markov lattice model with an efficient learning algorithm tailored for specific MRF subsets in computer vision, improving speed and robustness.

Findings

Faster learning algorithm with O(U log U) complexity

Effective modeling of specific MRF subsets in vision tasks

Robust performance on small datasets

Abstract

Markov random field (MRF) learning is intractable, and its approximation algorithms are computationally expensive. We target a small subset of MRF that is used frequently in computer vision. We characterize this subset with three concepts: Lattice, Homogeneity, and Inertia; and design a non-markov model as an alternative. Our goal is robust learning from small datasets. Our learning algorithm uses vector quantization and, at time complexity O(U log U) for a dataset of U pixels, is much faster than that of general-purpose MRF.