Model based audio sequence alignment

We formulate alignment of multiple audio sequences in a probabilistic framework. Our approach defines a generative model for time varying features extracted from audio clips that are recorded independently and asynchronously. We are able to handle missing data and multiple clips where no clip is covering the entire material. The matching is achieved via approximate Bayesian inference. Here, we illustrate a simulated tempering approach for sampling from the exact posterior density of the clip offsets. The simulation results on synthetic and real data suggest that the framework is able to handle difficult ambiguous scenarios or partial matchings.