Back to Search
Start Over
Blind Deinterleaving of Signals in Time Series with Self-attention Based Soft Min-cost Flow Learning
- Publication Year :
- 2020
-
Abstract
- We propose an end-to-end learning approach to address deinterleaving of patterns in time series, in particular, radar signals. We link signal clustering problem to min-cost flow as an equivalent problem once the proper costs exist. We formulate a bi-level optimization problem involving min-cost flow as a sub-problem to learn such costs from the supervised training data. We then approximate the lower level optimization problem by self-attention based neural networks and provide a trainable framework that clusters the patterns in the input as the distinct flows. We evaluate our method with extensive experiments on a large dataset with several challenging scenarios to show the efficiency.<br />Comment: 4 pages, 2 figures, 1 table
Details
- Database :
- arXiv
- Publication Type :
- Report
- Accession number :
- edsarx.2010.12972
- Document Type :
- Working Paper
- Full Text :
- https://doi.org/10.1109/ICASSP39728.2021.9415025