CITATION — REFERENCE ENTRY

time-domain · luo2019-conv-tasnet

Revision 5414605e-a9e2-4d0c-9ec4-95f958548e69 · 11/21/2026, 6:51:29 AM UTC
Claim ID
time-domain
Assertion
Conv-TasNet uses a learned linear encoder of the waveform and separates speakers by applying masks to the encoder output.
Quote
Conv-TasNet uses a linear encoder to generate a representation of the speech waveform optimized for separating individual speakers. Speaker separation is achieved by applying a set of weighting functions (masks) to the encoder output.
Quote language
en
Locator
page: 1 (Abstract)
Available in