chrissewell/whisperX
1
0
Fork 0
mirror of https://github.com/m-bain/whisperX.git synced 2025-07-01 18:17:27 -04:00
Code Issues Packages Projects Releases Wiki Activity

Pad language detection if audio is too short

Browse Source
This commit is contained in:
Simon
2023-04-30 18:34:18 +02:00
parent a903e57cf1
commit b9c8c5072b
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
whisperx/asr.py
View File

@ -251,7 +251,10 @@ class FasterWhisperPipeline(Pipeline):
def detect_language(self, audio: np.ndarray): def detect_language(self, audio: np.ndarray):
segment = log_mel_spectrogram(audio[: N_SAMPLES], padding=0) if audio.shape[0] < N_SAMPLES:
print("Warning: audio is shorter than 30s, language detection may be inaccurate.")
segment = log_mel_spectrogram(audio[: N_SAMPLES],
padding=0 if audio.shape[0] >= N_SAMPLES else N_SAMPLES - audio.shape[0])
encoder_output = self.model.encode(segment) encoder_output = self.model.encode(segment)
results = self.model.model.detect_language(encoder_output) results = self.model.model.detect_language(encoder_output)
language_token, language_probability = results[0][0] language_token, language_probability = results[0][0]