Merge pull request #4 from Jackwaterveg/cli_infer

revise
pull/1048/head
KP 3 years ago committed by GitHub
commit 2135aeb08d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

@ -228,10 +228,15 @@ class ASRExecutor(BaseExecutor):
audio = audio.mean(axis=1) audio = audio.mean(axis=1)
else: else:
audio = audio[:, 0] audio = audio[:, 0]
# pcm16 -> pcm 32
audio = audio.astype("float32") audio = audio.astype("float32")
bits = np.iinfo(np.int16).bits
audio = audio / (2**(bits - 1))
audio = librosa.resample(audio, audio_sample_rate, audio = librosa.resample(audio, audio_sample_rate,
self.sample_rate) self.sample_rate)
audio_sample_rate = self.sample_rate audio_sample_rate = self.sample_rate
# pcm32 -> pcm 16
audio = audio * (2**(bits - 1))
audio = np.round(audio).astype("int16") audio = np.round(audio).astype("int16")
else: else:
audio = audio[:, 0] audio = audio[:, 0]
@ -342,7 +347,7 @@ class ASRExecutor(BaseExecutor):
"The sample rate of the input file is not {}.\n \ "The sample rate of the input file is not {}.\n \
The program will resample the wav file to {}.\n \ The program will resample the wav file to {}.\n \
If the result does not meet your expectations\n \ If the result does not meet your expectations\n \
Please input the 16k 16bit 1 channel wav file. \ Please input the 16k 16 bit 1 channel wav file. \
" "
.format(self.sample_rate, self.sample_rate)) .format(self.sample_rate, self.sample_rate))
while (True): while (True):

Loading…
Cancel
Save