i had tried 8 wav files ,all are angry ,does it right?

[pinganzhang]

INFO:root:Final Emotion set:: {'Angry': 1.0, 'data': 'but on campus incident'}
{'Angry': 1.0, 'data': 'but on campus incident'

INFO:root:Final Emotion set:: {'Angry': 1.0, 'data': 'if one of the move'}
{'Angry': 1.0, 'data': 'if one of the move'}

do i need to train my own model?

[vyassu]

You might have to train your own model as the model uploaded is trained on very small dataset.

[pinganzhang]

Thank you @vyassu

[pinganzhang]

Did I train right? it seems doesn't change the results ?

WELCOME TO DEEPSENTIMENT

1. TEST using .wav file
2. TRAIN-TEST Your own model
   2
   Enter WAVE Filepath
   /home/pi/test.wav
   Enter Audio(WAVE File) test-train data filepath
   /home/pi/test.wav
   Enter Positive Train data filepath
   /home/pi/DeepSentiment/Code/StandAlone/Models
   Enter Negative Train data filepath
   /home/pi/DeepSentiment/Code/StandAlone/Models
   Enter Positive Test data filepath
   /home/pi/DeepSentiment/Code/StandAlone/Models
   Enter Negative Test data filepath
   /home/pi/DeepSentiment/Code/StandAlone/Models
   INFO:root:Loaded the Speech Recognition Module
   INFO:root:Saved the input signal into WAVE File
   INFO:root:Converted Speech To Text
   INFO:root:Training of Model Enabled
   INFO:root:Finished Loading SpeechNetSVMMulticlass Module
   INFO:root:Speech Model Training Started!!
   INFO:root:Finished Loading Test-Train data from file
   INFO:root:Emotion Data: {'Angry': 1.0}
   INFO:root:Retrieving the Training-Test Data for Speech2Text Model
   INFO:root:Retrieving Validation Data for Speech2Text Model
   INFO:root:Configuring the SVM Model
   INFO:root: Finished Loading the saved Speech2Text Model.
   INFO:root:Final Emotion set:: {'Angry': 1.0, 'data': "but that's not much and he seems not move"}
   {'Angry': 1.0, 'data': "but that's not much and he seems not move"}

[rishikksh20]

@vyassu Even I tried the same thing. Trained the model on 8 .wav files each having different emotions. However, it still classifies all subsequent .wav files as Angry. Where am I going wrong?