Multiple Instance Learning by Discriminative Training of Markov Networks

We introduce a graphical framework for multiple instance learning (MIL) based on Markov networks. This framework can be used to model the traditional MIL definition as well as more general MIL definitions. Different levels of ambiguity – the portion of positive instances in a bag – can be explored in weakly supervised data. To train these models, we propose a discriminative maxmargin learning algorithm leveraging efficient inference for cardinality-based cliques. The efficacy of the proposed framework is evaluated on a variety of data sets. Experimental results verify that encoding or learning the degree of ambiguity can improve classification performance.