facebookresearch/svoice
We provide a PyTorch implementation of the paper Voice Separation with an Unknown Number of Multiple Speakers In which, we present a new method for separating a mixed audio sequence, in which multiple voices speak simultaneously. The new method employs gated neural networks that are trained to separate the voices at multiple processing steps, while maintaining the speaker in each output channel fixed. A different model is trained for every number of possible speakers, and the model with the largest number of speakers is employed to select the actual number of speakers in a given sample. Our method greatly outperforms the current state of the art, which, as we show, is not competitive for more than two speakers.
PythonNOASSERTION
Issues
- 0
- 3
- 7
got soundfile.LibsndfileError: <unprintable LibsndfileError object> when execute train.py
#85 opened by carrud - 2
- 1
- 0
I need help with this problem: cannot import name 'get_ref_type' from 'omegaconf._utils'
#88 opened by Feavore - 2
- 0
Access is denied error
#92 opened by habib1402 - 0
- 0
- 2
the bee---------- sound?
#71 opened by cchoi1022 - 1
Evaluating
#67 opened by cchoi1022 - 1
how to solve cuda out of memory when execute train.py ? have tried to reduce batch size to 1 but problem still persist
#86 opened by carrud - 0
How to train using TPU?
#84 opened by mirosakr - 1
Pre trained model
#83 opened by jain11vaibhav - 11
Training for 10 speakers
#63 opened by muhammad-ahmed-ghani - 3
make_dataset.py is slow to generate samples
#74 opened by srdfjy - 2
fixed speaker
#75 opened by srdfjy - 1
- 0
Please help me
#82 opened by Deerzh - 0
RuntimeError: Offset past EOF
#81 opened by mirosakr - 3
AssertionError when training for 3 speakers
#54 opened by umerjamil16 - 0
Anyone got good results?
#79 opened by fabianoluzbr - 3
- 1
Training time using libri2mix
#66 opened by isakshamgoyal - 1
- 1
Guidelines for model training
#61 opened by arikhalperin - 1
- 2
Execute make_dataset err(ValueError: low >= high)
#73 opened by srdfjy - 2
Support for sm_86
#60 opened by arikhalperin - 0
Weight of Speaker ID loss
#70 opened by hsiaohan0827 - 1
max() arg is an empty sequence
#69 opened by joeoct93 - 0
Combining audio chunks
#68 opened by sealedtrainer - 1
- 0
convert model to onnx
#62 opened by Zed1014 - 1
separate error
#59 opened by linshuijin123 - 0
About the embedding model
#57 opened by YaFanYen - 3
Running into memory issue during training
#56 opened by arikhalperin - 1
Got ValueError when training with toy dataset
#55 opened by Yang-Xijie - 3
Scale-Invariant Source-to-Distortion Ratio (SI-SDR)
#52 opened by walidsst - 1
Run train.py script with torch 1.9.0
#53 opened by 7AM7 - 2
Speaker overlap between tr and cv
#50 opened by srdfjy - 4
The experimental result of the separation of two persons in svoice sisnr can reach "20.1". What are the audio numbers of tr, cv, and tt?
#49 opened by srdfjy - 4
- 2
- 0
ERROR:root:mic not in the room
#43 opened by srdfjy - 0
Do non-English training samples, synthetic samples and model configuration need special adjustments?
#45 opened by srdfjy - 0
Does svoice support "Early stopping"?
#46 opened by srdfjy - 2
- 2