Conv-TasNet
These files are modified according to
This Conv-TasNet model has been modified to target speaker separation .
Reference
Conv-Tesnet
"TasNet: Surpassing Ideal Time-Frequency Masking for Speech Separation".
Usage
Before running, you need to have mixture and target audio data.
Workflow
Workflow of egs/run.sh:
- Stage 1: Generating json files including wav path and duration
- Stage 2: Training
- Stage 3: Evaluate separation performance
- Stage 4: Separate speech using Conv-TasNet
How to resume training?
$ bash run.sh --continue_from <model-path>vad.py
filter out the silence in the signal