Spaces:
Running
Running
File size: 2,255 Bytes
c3e6538 f01c927 c3e6538 f01c927 c3e6538 4817b42 f01c927 c3e6538 4817b42 c3e6538 4817b42 c3e6538 4817b42 |
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 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
import torch
#from transformers import pipeline
#from transformers.pipelines.audio_utils import ffmpeg_read
from speechscore import SpeechScore
import gradio as gr
MODEL_NAME = "alibabasglab/speechscore"
BATCH_SIZE = 1
device = 0 if torch.cuda.is_available() else "cpu"
mySpeechScore = SpeechScore([
'PESQ','DNSMOS'
])
# Copied from https://github.com/openai/whisper/blob/c09a7ae299c4c34c5839a76380ae407e7d785914/whisper/utils.py#L50
def format_timestamp(seconds: float, always_include_hours: bool = False, decimal_marker: str = "."):
if seconds is not None:
milliseconds = round(seconds * 1000.0)
hours = milliseconds // 3_600_000
milliseconds -= hours * 3_600_000
minutes = milliseconds // 60_000
milliseconds -= minutes * 60_000
seconds = milliseconds // 1_000
milliseconds -= seconds * 1_000
hours_marker = f"{hours:02d}:" if always_include_hours or hours > 0 else ""
return f"{hours_marker}{minutes:02d}:{seconds:02d}{decimal_marker}{milliseconds:03d}"
else:
# we have a malformed timestamp so just return it as is
return seconds
def score(test_file, ref_file, dnsmos, pesq, return_timestamps):
scores = mySpeechScore(test_path=test_file, reference_path=ref_file, dnsmos=dnsmos, pesq=pesq, window=None, score_rate=16000, return_mean=False)
return scores
demo = gr.Blocks()
file_score = gr.Interface(
fn=score,
inputs=[
gr.Audio(sources=["upload"], label="test file", type="filepath"),
gr.Audio(sources=["upload"], label="reference file", type="filepath"),
#gr.Radio(["without reference", "with reference"], label="Task", info="choose non-instrusive or instrusive scoring"),
gr.Checkbox(default=False, label="DNSMOS"),
gr.Checkbox(default=False, label="PESQ"),
],
outputs="text",
#layout="horizontal",
#theme="huggingface",
title="Score speech from a file",
description=(
"Score audio inputs with the click of a button! Demo uses the"
" commonly used speech quality assessment methods for the audio files"
" of arbitrary length."
),
)
with demo:
gr.TabbedInterface([file_score], ["Score Audio File"])
demo.launch() |