sort INTEL NAT output
Browse files
app.py
CHANGED
@@ -82,7 +82,7 @@ def calc_mos(audio_path, ref):
|
|
82 |
wav_vad = torchaudio.functional.vad(wav, sample_rate=sr)
|
83 |
ppm = len(lst_phonemes) / (wav_vad.shape[-1] / sr) * 60
|
84 |
|
85 |
-
return AVA_MOS,
|
86 |
|
87 |
|
88 |
description ="""
|
@@ -100,9 +100,9 @@ iface = gr.Interface(
|
|
100 |
fn=calc_mos,
|
101 |
inputs=[gr.Audio(type='filepath', label="Audio to evaluate"),
|
102 |
gr.Textbox(placeholder="Input reference here (Don't keep this empty)", label="Reference")],
|
103 |
-
outputs=[gr.Textbox(placeholder="Naturalness Score
|
104 |
-
gr.Textbox(placeholder="Hypothesis", label="Hypothesis"),
|
105 |
gr.Textbox(placeholder="Intelligibility Score", label = "Intelligibility Score, range from 0 to 100, the higher the better"),
|
|
|
106 |
gr.Textbox(placeholder="Predicted Phonemes", label="Predicted Phonemes"),
|
107 |
gr.Textbox(placeholder="Speaking Rate, Phonemes per minutes", label="PPM")],
|
108 |
title="Laronix's Voice Quality Checking System Demo",
|
|
|
82 |
wav_vad = torchaudio.functional.vad(wav, sample_rate=sr)
|
83 |
ppm = len(lst_phonemes) / (wav_vad.shape[-1] / sr) * 60
|
84 |
|
85 |
+
return AVA_MOS, INTELI_score, trans, phone_transcription, ppm
|
86 |
|
87 |
|
88 |
description ="""
|
|
|
100 |
fn=calc_mos,
|
101 |
inputs=[gr.Audio(type='filepath', label="Audio to evaluate"),
|
102 |
gr.Textbox(placeholder="Input reference here (Don't keep this empty)", label="Reference")],
|
103 |
+
outputs=[gr.Textbox(placeholder="Naturalness Score", label="Naturalness Score, ranged from 0 to 5, the higher the better."),
|
|
|
104 |
gr.Textbox(placeholder="Intelligibility Score", label = "Intelligibility Score, range from 0 to 100, the higher the better"),
|
105 |
+
gr.Textbox(placeholder="Hypothesis", label="Hypothesis"),
|
106 |
gr.Textbox(placeholder="Predicted Phonemes", label="Predicted Phonemes"),
|
107 |
gr.Textbox(placeholder="Speaking Rate, Phonemes per minutes", label="PPM")],
|
108 |
title="Laronix's Voice Quality Checking System Demo",
|