Speech recognition on raspberrry pi

Here we put those posts that don’t have specified Category.

The tutorial is very nicely made and practical one . I have gone through all the three episodes. The tutorial site address is : for reference .

https://www.digikey.com/en/maker/projects/tensorflow-lite-tutorial-part-2-speech-recognition-model-training/d8d04a2b60a442cf8c3fa5c0dd2a292b

However the tutorial explains about single wake word “stop” if i like to add an other one like ‘marvin’, or yes, or no, ‘down’, up etc

github reference : https://github.com/ShawnHymel/tflite-speech-recognition/blob/master/02-speech-commands-mfcc-classifier.ipynb

exact para :

Settings

feature_sets_path = ‘C:\PATH\TO\feature_sets_directory’
feature_sets_filename = ‘all_targets_mfcc_sets.npz’
model_filename = ‘wake_word_stop_model.h5’
wake_word = ‘stop’

&

Convert ground truth arrays to one wake word (1) and ‘other’ (0)

wake_word_index = all_targets.index(wake_word)
y_train = np.equal(y_train, wake_word_index).astype(‘float64’)
y_val = np.equal(y_val, wake_word_index).astype(‘float64’)
y_test = np.equal(y_test, wake_word_index).astype(‘float64’)

where to add another wake word … please reply…

regards,hem.