Beta 1


Title Probabilistic Blind Source Separation
Author Olsson, Rasmus Kongsgaard (Intelligent Signal Processing, Informatics and Mathematical Modelling, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Master's thesis
Year 2004
Abstract This thesis focuses on Blind source separation (BSS), which is the problem of finding hidden source signals in observed mixtures given no or little knowledge about the sources and the mixtures. Based on the well-performing, yet heuristically based, algorithm of Parra and Spence, 2000, a probabilistic model is formulated for the BSS problem. A time-domain EM algorithm KaBSS is derived which estimates the source signals, the associated second-order statistics, the mixing filters and the observation noise covariance matrix. In line with the literature, it is found that the estimated quantities are unique within the model only if the sources can be assumed non-stationary and contain sufficient time-variation. Furthermore, the statistical framework is exploited in order to assess the correct model order: the number of sources within the mixture can be determined using the socalled Bayes Information Criterion (BIC). Monte Carlo simulations as well as experimental results for mixtures of speech signals are documented and compared to results obtained by the algorithm of Parra and Spence. In Danish: Denne afhandlings emne er blind signalseparation (BSS), der drejer sig om at estimere skjulte kildesignaler i observerede blandinger på basis af ringe eller ingen viden om kildesignaler og blandinger. En probabilistisk model for BSS-problemet formuleres med afsæt i Parra og Spences (2000) højtydende, men heuristisk funderede algoritme. På baggrund af modellen udledes KaBSS , en EM-algoritme, der estimerer kildesignalerne og deres 2. ordensstatistik, blandingsfiltrene og observationsstøjens kovarians. I overensstemmelse med litteraturen findes det, at de estimerede størrelser kun er unikke indenfor modellen, hvis en antagelse om kildernes ikke-stationaritet er rimelig, og hvis kilderne er tilstrækkelig tidsvariante. Ydermere udnyttes den statistiske ramme til at vurdere den korrekte modelorden: Antallet af kilder i blandingen fastslås ved at benytte det såkaldte Bayes Information Criterion (BIC). Såvel Monte Carlo simulationer som eksperimentelle resultater for blandinger af talesignaler dokumenteres og sammenlignes med resultater, opn°aet via Parra og Spences algoritme.
Note Supervised by Professor Lars Kai Hansen
Keywords Blind source separation; Independent component analysis; non-stationary sources; EM
Fulltext
Original Postscript imm3027.ps (3.19 MB)
Derived PDF imm3027.pdf (1.21 MB)
Admin Creation date: 2006-06-22    Update date: 2007-09-11    Source: dtu    ID: 154788    Original MXD