Unified auditory functions based on Bayesian topic model

摘要: Existing auditory functions for robots such as sound source localization and separation have been implemented in a cascaded framework whose overall performance may be degraded by any failure its subsystems. These approaches often require careful environment-dependent tuning each subsystems to achieve better performance. This paper presents unified where the whole system is integrated Bayesian topic model. method improves both with common configuration under various environments iterative inference using Gibbs sampling. Experimental results from three of different reverberation times confirm that our outperforms state-of-the-art methods, especially reverberant environments, shows comparable existing robot audition system.