gauravwaghmare / speaker-identification Goto Github PK

Hi @GauravWaghmare,

I have gone through your code.I didn't find the data which is required to run this code.
Can you please send me the links of Data-set or at-least please specify the format of sample Data-set applicable for this code.

Thanks and Regards

how to detect a new speaker

[root@aio Speaker-Identification]# python runMode_2.py train direc/
/root/anaconda2/lib/python2.7/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type.
from ._conv import register_converters as _register_converters
Using TensorFlow backend.
Segmentation fault (core dumped)

please give me suggestion?

I installed scikits.talkbox on python 2 and it's installed normally .The problem that i have to use python 3 because tensorflow is compatible only with python 3 on windows .Here some comments on the command prompt when i try to install scikits.talkbox.
For example:
*Running setup.py bdist_wheel for scikits.talkbox ... error
*No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils
*LINK : error LNK2001: unresolved external symbol PyInit__lpc
*build\temp.win-amd64-3.6\Release\scikits\talkbox\linpred\src_lpc.cp36-win_amd64.lib : fatal error LNK1120: 1 unresolved externals
*\scikits\talkbox\linpred_lpc.cp36-win_amd64.pyd /IMPLIB:build\temp.win-amd64-3.6\Release\scikits\talkbox\linpred\src_lpc.cp36-win_amd64.lib" failed with exit status 1120
*Failed building wheel for scikits.talkbox
*failed with error code 1 in C:\Users\DALIKA~1\AppData\Local\Temp\pip-build-sq6cbxni\scikits.talkbox
how can i resolve this problems please . Kinds regards

I am trying to run your code and I keep getting errors of deprecated libraries. Please include the versions of each python module you are using aka which Keras version? which Tensorflow version? etc.

how many datasets do we need to train?
such as ,how much audio per person?

CAN you tell me which paper this work is based on?

Hi @GauravWaghmare ,
I have gone through your code and train the model for sample of 5 speakers with 2 utterances for each speaker it identifies the speaker correctly. Can you please tell me the max number of speakers it identifies correctly and should we do transfer learning for this to add a new speaker or batch of speakers.

Thank you.

It does not support python 3.

I have about 340 speakers' speech,but after training the accuracy is very low