Spaces:
Runtime error
Runtime error
File size: 1,482 Bytes
91295c2 c5abae9 91295c2 59e6133 c5abae9 91295c2 c5abae9 530fef9 0e8cfdb c5abae9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 |
import gradio as gr
from pyannote.audio import Pipeline
from transformers import pipeline
asr = pipeline(
"automatic-speech-recognition",
model="facebook/wav2vec2-large-960h-lv60-self",
feature_extractor="facebook/wav2vec2-large-960h-lv60-self",
)
speaker_segmentation = Pipeline.from_pretrained("pyannote/speaker-segmentation")
def segmentation(audio):
speaker_output = speaker_segmentation(audio)
text_output = asr(audio,return_timestamps="word")
full_text = text_output['text'].lower()
chunks = text_output['chunks']
diarized_output = ""
i = 0
for turn, _, speaker in speaker_output.itertracks(yield_label=True):
diarized = ""
while i < len(chunks) and chunks[i]['timestamp'][1] <= turn.end:
diarized += chunks[i]['text'].lower() + ' '
i += 1
if diarized != "":
diarized_output += "{}: ''{}'' from {:.3f}-{:.3f}\n".format(speaker,diarized,turn.start,turn.end)
return diarized_output, full_text
inputs = gr.inputs.Audio(source="upload", type="filepath", label="Upload your audio file here:")
outputs = [gr.outputs.Textbox(type="auto", label="Diarized Output"),
gr.outputs.Textbox(type="auto",label="Full Text")]
examples = [["TestAudio1.wav"],]
app = gr.Interface(fn=segmentation,
inputs=inputs,
outputs=outputs,
examples=examples,
allow_flagging=False)
app.launch() |