Conv-TasNet Libri1Mix (vokra)

A Conv-TasNet speech enhancement model trained on the Libri1Mix enh_single task at 16 kHz, capable of separating and cleaning speech from noisy single-channel audio mixtures entirely in the time domain. It is derived from the Asteroid librimix recipe and closely mirrors the JorisCos/ConvTasNet_Libri1Mix_enhsingle_16k reference artifact. Source: supabase.