A common approach to blind source separation is to use independent component analysis. However when dealing with realistic convolutive audio and speech mixtures, processing in the frequency domain at each frequency bin is required. As a result this introduces the permutation problem, inherent in independent component analysis, across the frequency bins. Independent vector analysis directly addresses this issue by modeling the dependencies between frequency bins, namely making use of a source prior. An alternative source prior for real-time (online) natural gradient independent vector analysis is proposed. A Student's t probability density function is known to be more suited for speech sources, due to its heavier tails, and is incorporated into a real-time version of natural gradient independent vector analysis. In addition, the importance of the degrees of freedom parameter within the Student's t distribution is highlighted. The final algorithm is realized as a real-time embedded application on a floating point Texas Instruments digital signal processor platform, where simulated recordings from a reverberant room are used for testing. Results are shown to be better than with the original (super-Gaussian) source prior.
International audienceWe consider the design of a pair of time-domain filters to achieve target signal cancelation in a multi-source envi-ronment. The problem is formulated as a minimization of a sum squared error cost function with respect to the pair of finite impulse response cancelation filters. Direct mini-mization is achieved through an alternating gradient descent based method, whereas a novel method based on the method of principal angles is proposed which exploits the singular value decomposition. Simulation studies show that the gra-dient descent method suffers from slow convergence but this is overcome by the method based on principal angles which also achieves a lower cost than the gradient descent approach. The cancelation filters are then combined with an adaptive fil-tering scheme to address a video-informed audio source sep-aration problem and preliminary results suggest good perfor-mance in terms of objective measures
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.