Learning Time Series Segmentation Models from Temporally Imprecise Labels

Abstract

This paper considers the problem of learning time series segmentation models when the labeled data is subject to temporal uncertainty or noise. Our approach augments the semi-Markov conditional random field (semi-CRF) model with a probabilistic model of the label observation process. This augmentation allows us to estimate the parameters of the semi-CRF from timestamps corresponding roughly to the occurrence of segment transitions. We show how exact marginal inference can be performed in polynomial time enabling learning based on marginal likelihood maximization. Our experiments on two activity detection problems show that the proposed approach can learn models from temporally imprecise labels, and can successfully refine imprecise segmentations through posterior inference. Finally, we show how inference complexity can be reduced by a factor of 40 using static and model-based pruning of the inference dynamic program.