Partial Membership Latent Dirichlet Allocation

TL;DR

This paper introduces PM-LDA, a novel topic model that enables both crisp and soft image segmentation by allowing partial memberships, addressing limitations of traditional models in handling transition regions.

Contribution

The paper proposes the PM-LDA model and algorithms, enabling partial memberships in topic modeling for improved image segmentation over existing methods.

Findings

PM-LDA produces both crisp and soft segmentations.

Experimental results show PM-LDA outperforms existing models.

Applicable to natural and SONAR image datasets.

Abstract

Topic models (e.g., pLSA, LDA, SLDA) have been widely used for segmenting imagery. These models are confined to crisp segmentation. Yet, there are many images in which some regions cannot be assigned a crisp label (e.g., transition regions between a foggy sky and the ground or between sand and water at a beach). In these cases, a visual word is best represented with partial memberships across multiple topics. To address this, we present a partial membership latent Dirichlet allocation (PM-LDA) model and associated parameter estimation algorithms. Experimental results on two natural image datasets and one SONAR image dataset show that PM-LDA can produce both crisp and soft semantic image segmentations; a capability existing methods do not have.