Update app.py
Browse files
app.py
CHANGED
@@ -17,7 +17,12 @@ def ASR(audio):
|
|
17 |
temp_audio_file.flush()
|
18 |
# Load the audio file using torchaudio
|
19 |
waveform, sample_rate = torchaudio.load(temp_audio_file.name)
|
20 |
-
|
|
|
|
|
|
|
|
|
|
|
21 |
# Convert the PyTorch tensor to a NumPy ndarray
|
22 |
audio_array = waveform.numpy()
|
23 |
#inputs = processor(audio_array, sampling_rate=16_000)
|
|
|
17 |
temp_audio_file.flush()
|
18 |
# Load the audio file using torchaudio
|
19 |
waveform, sample_rate = torchaudio.load(temp_audio_file.name)
|
20 |
+
# Resample the audio to 16kHz
|
21 |
+
resampler = torchaudio.transforms.Resample(sample_rate, 16000)
|
22 |
+
waveform = resampler(waveform)
|
23 |
+
# Convert the audio to a single channel
|
24 |
+
downmix_mono = torchaudio.transforms.DownmixMono()
|
25 |
+
waveform = downmix_mono(waveform)
|
26 |
# Convert the PyTorch tensor to a NumPy ndarray
|
27 |
audio_array = waveform.numpy()
|
28 |
#inputs = processor(audio_array, sampling_rate=16_000)
|