shahules786 / mayavoz Goto Github PK

Hi,

Great repo! Is this work on real-time?

Thanks

Thanks for this nice package. I'm curious if there are any models that have been trained with a sampling rate about 16 kHz?

Exception has occurred: ParameterError
Mono data must have shape (samples,). Received shape=(1, 320000)

Hi,

I love what you are doing but I hit issues with 3.9 on:

Building wheels for collected packages: pesq, shap
Building wheel for pesq (setup.py) ... error
error: subprocess-exited-with-error

× python setup.py bdist_wheel did not run successfully.
│ exit code: 1
╰─> [28 lines of output]
Traceback (most recent call last):
File "", line 2, in
File "", line 34, in
File "/private/var/folders/t1/8wysbvsx2h50nm_26pqnq2780000gn/T/pip-install-_0o0kg7n/pesq_574d32d341a945a6b6fad78be2b203af/setup.py", line 33, in
setup(
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/init.py", line 152, in setup
_install_setup_requires(attrs)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/init.py", line 147, in _install_setup_requires
dist.fetch_build_eggs(dist.setup_requires)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/dist.py", line 806, in fetch_build_eggs
resolved_dists = pkg_resources.working_set.resolve(
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/pkg_resources/init.py", line 766, in resolve
dist = best[req.key] = env.best_match(
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/pkg_resources/init.py", line 1051, in best_match
return self.obtain(req, installer)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/pkg_resources/init.py", line 1063, in obtain
return installer(requirement)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/dist.py", line 877, in fetch_build_egg
return fetch_build_egg(self, req)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/installer.py", line 80, in fetch_build_egg
wheel.install_as_egg(dist_location)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/wheel.py", line 95, in install_as_egg
self._install_as_egg(destination_eggdir, zf)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/wheel.py", line 103, in _install_as_egg
self._convert_metadata(zf, destination_eggdir, dist_info, egg_info)
File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/setuptools/wheel.py", line 124, in _convert_metadata
os.mkdir(destination_eggdir)
FileExistsError: [Errno 17] File exists: '/private/var/folders/t1/8wysbvsx2h50nm_26pqnq2780000gn/T/pip-install-_0o0kg7n/pesq_574d32d341a945a6b6fad78be2b203af/.eggs/numpy-1.24.0-py3.9-macosx-10.9-universal2.egg'

so I upgraded to 3.11 and hit:

❯ /usr/local/bin/pip3 install mayavoz
Collecting mayavoz
Using cached mayavoz-0.0.1-py2.py3-none-any.whl (38 kB)
Collecting boto3>=1.24.86
Using cached boto3-1.26.36-py3-none-any.whl (132 kB)
Collecting huggingface-hub>=0.10.0
Using cached huggingface_hub-0.11.1-py3-none-any.whl (182 kB)
Collecting hydra-core>=1.2.0
Using cached hydra_core-1.3.1-py3-none-any.whl (154 kB)
Collecting joblib>=1.2.0
Using cached joblib-1.2.0-py3-none-any.whl (297 kB)
Collecting librosa>=0.9.2
Using cached librosa-0.9.2-py3-none-any.whl (214 kB)
Collecting mlflow>=1.28.0
Using cached mlflow-2.1.0-py3-none-any.whl (16.7 MB)
Collecting numpy>=1.23.3
Downloading numpy-1.24.0-cp311-cp311-macosx_11_0_arm64.whl (13.8 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 13.8/13.8 MB 5.1 MB/s eta 0:00:00
Collecting pesq==0.0.4
Using cached pesq-0.0.4.tar.gz (38 kB)
Preparing metadata (setup.py) ... done
Collecting protobuf>=3.19.6
Using cached protobuf-4.21.12-cp37-abi3-macosx_10_9_universal2.whl (486 kB)
Collecting pystoi==0.3.3
Using cached pystoi-0.3.3.tar.gz (7.0 kB)
Preparing metadata (setup.py) ... done
Collecting pytest-lazy-fixture>=0.6.3
Using cached pytest_lazy_fixture-0.6.3-py3-none-any.whl (4.9 kB)
Collecting pytorch-lightning>=1.7.7
Using cached pytorch_lightning-1.8.6-py3-none-any.whl (800 kB)
Collecting scikit-learn>=1.1.2
Downloading scikit_learn-1.2.0-cp311-cp311-macosx_12_0_arm64.whl (8.3 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 8.3/8.3 MB 5.8 MB/s eta 0:00:00
Collecting scipy>=1.9.1
Downloading scipy-1.9.3-cp311-cp311-macosx_12_0_arm64.whl (28.4 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 28.4/28.4 MB 6.5 MB/s eta 0:00:00
Collecting soundfile>=0.11.0
Using cached soundfile-0.11.0-py2.py3-none-macosx_10_9_arm64.macosx_11_0_arm64.whl (1.1 MB)
ERROR: Ignored the following versions that require a different python version: 1.21.2 Requires-Python >=3.7,<3.11; 1.21.3 Requires-Python >=3.7,<3.11; 1.21.4 Requires-Python >=3.7,<3.11; 1.21.5 Requires-Python >=3.7,<3.11; 1.21.6 Requires-Python >=3.7,<3.11; 1.6.2 Requires-Python >=3.7,<3.10; 1.6.3 Requires-Python >=3.7,<3.10; 1.7.0 Requires-Python >=3.7,<3.10; 1.7.0rc1 Requires-Python >=3.7,<3.10; 1.7.0rc2 Requires-Python >=3.7,<3.10; 1.7.1 Requires-Python >=3.7,<3.10; 1.7.2 Requires-Python >=3.7,<3.11; 1.7.3 Requires-Python >=3.7,<3.11; 1.8.0 Requires-Python >=3.8,<3.11; 1.8.0rc1 Requires-Python >=3.8,<3.11; 1.8.0rc2 Requires-Python >=3.8,<3.11; 1.8.0rc3 Requires-Python >=3.8,<3.11; 1.8.0rc4 Requires-Python >=3.8,<3.11; 1.8.1 Requires-Python >=3.8,<3.11
ERROR: Could not find a version that satisfies the requirement torch>=1.12.1 (from mayavoz) (from versions: none)
ERROR: No matching distribution found for torch>=1.12.1

In the conv.py file, should lines 73 and 132 be:

imag = real_imag + imag_real

rather than

imag = real_imag - imag_real

This would make them consistent with equation 1 in the DCCRN paper and also consistent with how things are defined in the rnn.py file for the complex LSTM.

While writing output mayavoz currently uses the sampling rate of the pre-trained model used as the data sampling rate. This can cause issues for users who want to preserve the original input sampling rate in the output.

Add more/better pre-trained models to mayavoz.

At line 37 in CONTRIBUTING.md,

Instead of 'pytest' , it spelled as 'pyest'.

I gave this a try with a one-minute sample of a speaker in a noisy room. It's very intelligible to start, but after running through the filter I just got pops and squeaks. I then tried normalizing it first, which produced an intelligible result, but very artificial and generally not as good as the original.

This is not a complaint! More a query to see if you are interested to have the before and after samples.

I'm using the technique in the README, along with the save_output flag. My input file is 44k mono WAV.

You probably should rename the license owner ;-)

And if there is code which is inspired or taken from Asteroid, giving it credits would be very cool.

Best of luck to mayavoz !

Hi,
I used your pretrained DCCRN, but there are some problem

First, it requires 2 seconds longer signal
And, the output value are NaNs
I tried with sample audio in tests/data/vctk.
And my environment is titanX with cuda 11.4 with torch==1.12.1+cu113

waveunet and demucs are working well

Hi,
It seems like that torch lightning 2.0.0 did't support "pytorch_lightning.utilities.cloud_io"
Can you update code or requirements??

Add this blog to the readme to educate users about SE techniques.

The example fails with:

ModuleNotFoundError: No module named 'pytorch_lightning.utilities.cloud_io

I did some grepping around and spotted that pytorch_lightning now imports that from a separate package. Example:

from lightning_fabric.utilities.cloud_io import get_filesystem

I'm attempting to use Python 3.11 on Ubuntu, in a Python virtual environment so as to allow use of pip.

Hope this is helpful! Thanks for open-sourcing this.

Currently mayavoz only reads .wav files. Extend support for .mp3 files.

I am using a virtual environment created with virtualenv.

I am using version:
Python 3.8.10
pip 20.0.2
Ubuntu 20.04 Windows Subsystem for Linux

while using a virtual environment, I run:

pip install -r requirements.txt

While this is running, I receive this error message

ERROR: numba 0.56.4 has requirement numpy<1.24,>=1.18, but you'll have numpy 1.24.0 which is incompatible.
ERROR: mlflow 2.0.1 has requirement packaging<22, but you'll have packaging 22.0 which is incompatible.
ERROR: tensorboardx 2.5.1 has requirement protobuf<=3.20.1,>=3.8.0, but you'll have protobuf 4.21.12 which is incompatible.

Just gave a try with this example :

from mayavoz.models import Mayamodel

model = Mayamodel.from_pretrained("shahules786/mayavoz-waveunet-valentini-28spk")
model.enhance("TRK010.WAV")

From command-line, got this error:

ImportError: cannot import name 'Mayamodel' from partially initialized module 'mayavoz.models' (most likely due to a circular import)

Tried from mayavoz.models.model import Mayamodel, got

ImportError: cannot import name 'move_data_to_device' from partially initialized module 'lightning_lite.utilities' (most likely due to a circular import)