prakashpandey9 / IsolatedSpeechRecognition

A python implementation of isolated word recognition using Hidden Markov Model
41 stars 21 forks source link

I am getting errors in executing the code. #3

Open Firedope opened 5 years ago

Firedope commented 5 years ago
  1. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\newSR.py", line 449, in ps_test = [m.transform(observation) for m in ms]
  2. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\newSR.py", line 449, in ps_test = [m.transform(observation) for m in ms]
  3. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\newSR.py", line 276, in transform B = self._state_likelihood(obs[n, :, :])
  4. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\newSR.py", line 167, in _state_likelihood obs.T, mean=self.mu[:, s].T, cov=self.covs[:, :, s].T)
  5. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\lib\site-packages\scipy\stats_multivariate.py", line 522, in pdf out = np.exp(self._logpdf(x, mean, psd.U, psd.log_pdet, psd.rank))
  6. File "C:\Users\Kanika Bansalwal\AppData\Local\Programs\Python\Python36\lib\site-packages\scipy\stats_multivariate.py", line 469, in _logpdf dev = x - mean ValueError: operands could not be broadcast together with shapes (7338,10) (6,)
Firedope commented 5 years ago

Please anyone can check this..?