Learning Structured Outputs from Partial Labels using Forest Ensemble

TL;DR

This paper introduces AdaBoost.MRF, an efficient boosting algorithm for learning structured outputs from partial labels, leveraging the superimposition of trees to handle complex models with guaranteed convergence.

Contribution

The paper presents a novel boosting algorithm that handles partial labels for structured output learning using tree-based models, ensuring convergence and practical applicability.

Findings

Successfully applied to indoor video surveillance data

Handles partial labels effectively in structured output learning

Guarantees convergence due to tree-based model structure

Abstract

Learning structured outputs with general structures is computationally challenging, except for tree-structured models. Thus we propose an efficient boosting-based algorithm AdaBoost.MRF for this task. The idea is based on the realization that a graph is a superimposition of trees. Different from most existing work, our algorithm can handle partial labelling, and thus is particularly attractive in practice where reliable labels are often sparsely observed. In addition, our method works exclusively on trees and thus is guaranteed to converge. We apply the AdaBoost.MRF algorithm to an indoor video surveillance scenario, where activities are modelled at multiple levels.