Update app.py
Browse files
app.py
CHANGED
@@ -21,8 +21,14 @@ def translate_speech(audio_data_tuple):
|
|
21 |
# Print the shape and type of the audio data
|
22 |
print(f"Audio data type: {type(audio_data)}, Audio data shape: {audio_data.shape}")
|
23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
# Use the speech recognition pipeline to transcribe the audio
|
25 |
-
output = pipe(
|
26 |
|
27 |
print(f"Output: {output}") # Print the output to see what it contains
|
28 |
|
|
|
21 |
# Print the shape and type of the audio data
|
22 |
print(f"Audio data type: {type(audio_data)}, Audio data shape: {audio_data.shape}")
|
23 |
|
24 |
+
# Normalize the audio data to the range [-1, 1]
|
25 |
+
audio_data_normalized = audio_data / np.iinfo(audio_data.dtype).max
|
26 |
+
|
27 |
+
# Convert the normalized audio data to float64
|
28 |
+
audio_data_float64 = audio_data_normalized.astype(np.float64)
|
29 |
+
|
30 |
# Use the speech recognition pipeline to transcribe the audio
|
31 |
+
output = pipe(audio_data_float64)
|
32 |
|
33 |
print(f"Output: {output}") # Print the output to see what it contains
|
34 |
|