Sparse Audio Inpainting: A Dictionary Learning Technique to Improve Its Performance

Tauboeck G, Rajbamshi S, Balazs P. Sparse Audio Inpainting: A Dictionary Learning Technique to Improve Its Performance. In: AES Convention 149. Audio Engineering Society; 2020. Paper 10402. Available from: https://aes.org/publications/elibrary-page/?id=20939

@inproceedings{Tauboeck2020_20939,
  author = {Tauboeck, Georg and Rajbamshi, Shristi and Balazs, Peter},
  title = {{Sparse Audio Inpainting: A Dictionary Learning Technique to Improve Its Performance}},
  booktitle = {AES Convention 149},
  note = {Paper 10402},
  year = {2020},
  month = oct,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=20939}
}

TY  - CPAPER
TI  - Sparse Audio Inpainting: A Dictionary Learning Technique to Improve Its Performance
AU  - Tauboeck, Georg
AU  - Rajbamshi, Shristi
AU  - Balazs, Peter
T2  - AES Convention 149
M1  - Paper 10402
PY  - 2020
DA  - 2020/10/06
UR  - https://aes.org/publications/elibrary-page/?id=20939
PB  - Audio Engineering Society
LA  - en
AB  - The objective of audio inpainting is to fill a gap in a signal, either to be meaningful or even to reconstruct the original signal. We propose a novel approach applying sparse modeling in the time-frequency (TF) domain. In particular, we develop a dictionary learning technique which deforms a given Gabor frame such that the sparsity of the analysis coefficients of the resulting frame is maximized. A suitable modification of the SParse Audio Inpainter (SPAIN) allows to exploit the obtained sparsity gain and, hence, to benefit from the learned dictionary. Our experiments demonstrate that our methods outperforms several state-of-the-art audio inpainting techniques in terms of signal-to-noise ratio (SNR) and objective difference grade (ODG).
ER  -