Lam, MWY;
Wang, J;
Su, D;
Yuy, D;
(2021)
Sandglasset: A Light Multi-Granularity Self-Attentive Network for Time-Domain Speech Separation.
In:
Proceedings of the ICASSP 2021 - IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
(pp. pp. 5759-5763).
IEEE
Preview |
Text
2103.00819.pdf - Accepted Version Download (760kB) | Preview |
Abstract
One of the leading single-channel speech separation (SS) models is based on a TasNet with a dual-path segmentation technique, where the size of each segment remains unchanged throughout all layers. In contrast, our key finding is that multi-granularity features are essential for enhancing contextual modeling and computational efficiency. We introduce a self-attentive network with a novel sandglass-shape, namely Sandglasset, which advances the state-of-the-art (SOTA) SS performance at significantly smaller model size and computational cost. Forward along each block inside Sandglasset, the temporal granularity of the features gradually becomes coarser until reaching half of the network blocks, and then successively turns finer towards the raw signal level. We also unfold that residual connections between features with the same granularity are critical for preserving information after passing through the bottleneck layer. Experiments show our Sandglasset with only 2.3M parameters has achieved the best results on two benchmark SS datasets - WSJ0-2mix and WSJ0- 3mix, where the SI-SNRi scores have been improved by absolute 0.6 dB and 2.4 dB, respectively, comparing to the prior SOTA results.
| Type: | Proceedings paper |
|---|---|
| Title: | Sandglasset: A Light Multi-Granularity Self-Attentive Network for Time-Domain Speech Separation |
| Event: | ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) |
| Location: | Toronto, ON, Canada |
| Dates: | 6th-11th June 2021 |
| ISBN-13: | 978-1-7281-7605-5 |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1109/ICASSP39728.2021.9413837 |
| Publisher version: | https://doi.org/10.1109/ICASSP39728.2021.9413837 |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher's terms and conditions. |
| Keywords: | Speech separation, multi-granularity, selfattentive network, single-channel |
| UCL classification: | UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science UCL > Provost and Vice Provost Offices > UCL BEAMS UCL |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10154107 |
Loading...
Loading...Loading...Loading...Archive Staff Only
 |
View Item |
