Supervised and Semi-supervised Separation of Sounds from Single-Channel Mixtures
In this paper we describe a methodology for model-based single channel separation of sounds. We present a sparse latent variable model that can learn sounds based on their distribution of time/ frequency energy. This model can then be used to extract known types of sounds from mixtures in two scenarios. One being the case where all sound types in the mixture are known, and the other being being the case where only the target or the interference models are known. The model we propose has close ties to non-negative decompositions and latent variable models commonly used for semantic analysis.
Unable to display preview. Download preview PDF.
- 1.Casey, M., Westner, A.: Separation of Mixed Audio Sources by Independent Subspace Analysis. In: proceedings ICMC (2000)Google Scholar
- 2.Roweis, S.T.: One Microphone Source Separation. In: NIPS (2000)Google Scholar
- 3.Benaroya, L., McDonagh, L., Bimbot, F., Gribonval, R.: Non negative sparse representation for Wiener based source separation with a single sensor. In: proceedings of the ICASSP (2003)Google Scholar
- 4.Vincent, E., Rodet, X.: Music transcription with ISA and HMM. In: Puntonet, C.G., Prieto, A.G. (eds.) ICA 2004. LNCS, vol. 3195, Springer, Heidelberg (2004)Google Scholar
- 5.Hofmann, T.: Probabilistic Latent Semantic Indexing. In: proceedings SIGIR 1999 (1999)Google Scholar
- 7.Shashanka, M.V.S.: A Unified Probabilistic Approach to Modeling and Separating Single-Channel Acoustic Sources, Ph.D. Thesis, Department of Cognitive and Neural Systems. Boston University, Boston (2007)Google Scholar
- 8.Lee, D.D, Seung, H.S.: Algorithms for Non-negative Matrix Factorization. In: NIPS (2001)Google Scholar