Exploiting Contextual Knowledge for Hybrid Classification of Visual Objects

Abstract

We consider the problem of classifying visual objects in a scene by exploiting the semantic context. For this task, we define hybrid classifiers (HC) that combine local classifiers with context constraints, and can be applied to collective classification problems (CCPs) in general. Context constraints are represented by weighted ASP constraints using object relations. To integrate probabilistic information provided by the classifier and the context, we embed our encoding in the formalism LP M LN , and show that an optimal labeling can be efficiently obtained from the corresponding LP M LN program by employing an ordinary ASP solver. Moreover, we describe a methodology for constructing an HC for a CCP, and present experimental results of applying an HC for object classification in indoor and outdoor scenes, which exhibit significant improvements in terms of accuracy compared to using only a local classifier.

Integrated Algorithms for HEX-Programs and Applications in Machine Learning

“…We approached the Collective Classification problem in ASP by defining Hybrid Classifiers (HC) that combine a local classifier, which predicts the probability of each local label based on object features, with context constraints (weighted ASP constraints) using object relations [9]. At this, external atoms of HEX can be used to interface an ontology reasoner, a spatial reasoning calculus as well as the local classifier directly from within the encoding.…”

On the Strong Equivalences of LPMLN Programs

“…Besides, several inference tasks are introduced to LP MLN such as computing marginal probability distribution of beliefs, computing most probable belief sets etc., which makes LP MLN suitable for knowledge reasoning in the context that contains uncertain and inconsistent data. For example, Eiter and Kaminski [5] used LP MLN in the tasks of classifying visual objects, and some unpublished work tried to use LP MLN as the bridge between text and logical knowledge bases.…”