[Guide] Trigger word - Try your own example!

trigger_word.py

# Set the filename
your_filename = "audio_examples/1040676.wav"
clip = AudioSegment.from_mp3("audio_examples/1040676.mp3")

# Add whitenoise to end of shorter samples
from pydub.generators import WhiteNoise
noise_bg = WhiteNoise().to_audio_segment(duration=clip.duration_seconds*1000.0, volume=-50)
noise_end = WhiteNoise().to_audio_segment(duration=max(10000.0-clip.duration_seconds*1000.0,0), volume=-50)
clip = clip.overlay(noise_bg) + noise_end

# Convert to wav
clip.export(your_filename, format="wav")

# Predict
preprocess_audio(your_filename)
chime_threshold = 0.8
prediction = detect_triggerword(your_filename)
chime_on_activate(your_filename, prediction, chime_threshold)
IPython.display.Audio("./chime_output.wav")