mirror of
https://github.com/k2-fsa/sherpa-onnx.git
synced 2026-11-10 21:11:03 +08:00
106 lines
3.9 KiB
Python
Executable File
106 lines
3.9 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
|
|
"""
|
|
This file shows how to use the speech enhancement API with DPDFNet.
|
|
|
|
Please download DPDFNet models from the sherpa-onnx GitHub release
|
|
or the official Hugging Face hub:
|
|
https://github.com/k2-fsa/sherpa-onnx/releases/tag/speech-enhancement-models
|
|
https://huggingface.co/Ceva-IP/DPDFNet
|
|
|
|
Example:
|
|
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/dpdfnet_baseline.onnx
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/dpdfnet2.onnx
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/dpdfnet4.onnx
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/dpdfnet8.onnx
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/dpdfnet2_48khz_hr.onnx
|
|
wget https://huggingface.co/Ceva-IP/DPDFNet/resolve/main/onnx/dpdfnet2_8khz.onnx
|
|
wget https://huggingface.co/Ceva-IP/DPDFNet/resolve/main/onnx/dpdfnet8_8khz.onnx
|
|
wget https://huggingface.co/Ceva-IP/DPDFNet/resolve/main/onnx/dpdfnet8_48khz_hr.onnx
|
|
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/speech-enhancement-models/speech_with_noise.wav
|
|
|
|
Use 16 kHz DPDFNet models such as `dpdfnet_baseline.onnx`, `dpdfnet2.onnx`,
|
|
`dpdfnet4.onnx`, or `dpdfnet8.onnx` for downstream ASR or speech recognition.
|
|
Use `dpdfnet2_8khz.onnx` or `dpdfnet8_8khz.onnx` for 8 kHz enhancement, and
|
|
`dpdfnet2_48khz_hr.onnx` or `dpdfnet8_48khz_hr.onnx` for 48 kHz output.
|
|
"""
|
|
|
|
import time
|
|
from pathlib import Path
|
|
from typing import Tuple
|
|
|
|
import numpy as np
|
|
import sherpa_onnx
|
|
import soundfile as sf
|
|
|
|
|
|
def create_speech_denoiser():
|
|
model_filename = "./dpdfnet_baseline.onnx"
|
|
if not Path(model_filename).is_file():
|
|
print(f"{model_filename} does not exist")
|
|
raise ValueError(
|
|
"Please first download a DPDFNet model from "
|
|
"the sherpa-onnx GitHub release or the official Hugging Face hub: "
|
|
"https://github.com/k2-fsa/sherpa-onnx/releases/tag/speech-enhancement-models or "
|
|
"https://huggingface.co/Ceva-IP/DPDFNet"
|
|
)
|
|
|
|
config = sherpa_onnx.OfflineSpeechDenoiserConfig(
|
|
model=sherpa_onnx.OfflineSpeechDenoiserModelConfig(
|
|
dpdfnet=sherpa_onnx.OfflineSpeechDenoiserDpdfNetModelConfig(
|
|
model=model_filename,
|
|
attenuation_limit_db=12.0,
|
|
),
|
|
debug=False,
|
|
num_threads=1,
|
|
provider="cpu",
|
|
),
|
|
)
|
|
if not config.validate():
|
|
print(config)
|
|
raise ValueError("Errors in config. Please check previous error logs")
|
|
return sherpa_onnx.OfflineSpeechDenoiser(config)
|
|
|
|
|
|
def load_audio(filename: str) -> Tuple[np.ndarray, int]:
|
|
data, sample_rate = sf.read(
|
|
filename,
|
|
always_2d=True,
|
|
dtype="float32",
|
|
)
|
|
data = data[:, 0] # use only the first channel
|
|
samples = np.ascontiguousarray(data)
|
|
return samples, sample_rate
|
|
|
|
|
|
def main():
|
|
sd = create_speech_denoiser()
|
|
test_wave = "./speech_with_noise.wav"
|
|
if not Path(test_wave).is_file():
|
|
raise ValueError(
|
|
f"{test_wave} does not exist. You can download it from "
|
|
"https://github.com/k2-fsa/sherpa-onnx/releases/tag/speech-enhancement-models"
|
|
)
|
|
|
|
samples, sample_rate = load_audio(test_wave)
|
|
|
|
start = time.time()
|
|
denoised = sd(samples, sample_rate)
|
|
end = time.time()
|
|
|
|
elapsed_seconds = end - start
|
|
audio_duration = len(samples) / sample_rate
|
|
real_time_factor = elapsed_seconds / audio_duration
|
|
|
|
output_filename = f"./enhanced_{denoised.sample_rate}.wav"
|
|
sf.write(output_filename, denoised.samples, denoised.sample_rate)
|
|
print(f"Saved to {output_filename}")
|
|
print(f"Elapsed seconds: {elapsed_seconds:.3f}")
|
|
print(f"Audio duration in seconds: {audio_duration:.3f}")
|
|
print(f"RTF: {elapsed_seconds:.3f}/{audio_duration:.3f} = {real_time_factor:.3f}")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|