Causal sparse decomposition of audio signals (bibtex)
by A. Charles, A.A. Kressner and C.J. Rozell
Abstract:
Recent results have shown the utility of sparse coding for audio signals [citation]. While current inference methods can decompose audio signals, they require whole signals and are therefor ill suited for realtime applications that require causal processing. We propose a neurally inspired, causal, sparse inference scheme based on the Locally Competitive Algorithm (LCA) (Rozell et al. 2008) over a temporal-spectral neighborhood. We demonstrate that this causal inference scheme can achieve lower sparsity levels and better signal fidelity than current filter and threshold approaches. Additionally, for some regimes, the sparsity level approaches those of Matching Pursuit while still maintaining signal integrity.
Reference:
Causal sparse decomposition of audio signalsA. Charles, A.A. Kressner and C.J. Rozell. January 2011.
Bibtex Entry:
@CONFERENCE{charles.11,
  author = 	 {Charles, A. and Kressner, A.A. and Rozell, C.J.},
  title = 	 {Causal sparse decomposition of audio signals},
  booktitle =	 {{Proceedings of the IEEE Digital Signal Processing (DSP) Workshop}},
  year =	 2011,
  month = {January},
  address =	 {Sedona, AZ},
abstract = {Recent results have shown the utility of sparse coding for audio signals [citation]. While current inference methods can decompose audio signals, they require whole signals and are therefor ill suited for realtime applications that require causal processing. We propose a neurally inspired, causal, sparse inference scheme based on the Locally Competitive Algorithm (LCA) (Rozell et al. 2008) over a temporal-spectral neighborhood. We demonstrate that this causal inference scheme can achieve lower sparsity levels and better signal fidelity than current filter and threshold approaches. Additionally, for some regimes, the sparsity level approaches those of Matching Pursuit while still maintaining signal integrity.}
}
Powered by bibtexbrowser