speech-selib 0.1.0

A simple speech enhancement library using ONNX models

SeLib

SeLib is a small speech-enhancement library for ONNX models. It provides a simple enhance() function, lightweight model wrappers, and utility functions for audio/noise experiments.

Install

Install the package directly from GitHub while the project is still small and moving quickly.

# Install SeLib and its Python dependencies.
pip install git+https://github.com/nipponjo/selib.git

Basic Use

The shortest path is selib.enhance(wave, model_id). Pass a mono NumPy waveform at the sample rate expected by the model.

import librosa
import selib

from selib.utils import NoiseGenerator, snr_mixer

# Load clean speech at the model sample rate.
wave_clean, sr = librosa.load("./data/clean_freesound_33711.wav", sr=48000)

# Create artificial noise and mix it with the clean speech at 5 dB SNR.
noise_gen = NoiseGenerator(sample_rate=sr)
noise = noise_gen.sample(n_samples=len(wave_clean))
wave_noisy = snr_mixer(wave_clean, noise, snr=5)

# Enhance with a registered model id.
# The ONNX model is downloaded and cached automatically on first use.
wave_enhanced = selib.enhance(wave_noisy, "deepfilternet3")

You can also save the enhanced waveform directly. The loaded model is cached by default, so repeated calls with the same model_id do not reload the ONNX session.

# Save as 24-bit PCM WAV while returning the enhanced NumPy array.
wave_enhanced = selib.enhance(
    wave_noisy,
    "deepfilternet3",    
    save_to="enhanced.wav",
    bits_per_sample=24,
)

Model Objects

For more control, load the model wrapper once and call its enhance() method directly.

from selib import load_model
from selib.models import DeepFilterNetOnnx

# load_model() chooses the correct wrapper from the registry metadata.
model: DeepFilterNetOnnx = load_model("deepfilternet3")

# DeepFilterNetOnnx supports extra options such as attenuation limiting.
wave_enhanced = model.enhance(wave_noisy, atten_lim_db=12)

Magnitude-mask models can also be used directly. They predict a mask, multiply it with the noisy magnitude spectrogram, and reuse the noisy phase.

from selib import MagnitudeMaskModel

# ul_unas_16k is a 16 kHz magnitude-mask model.
mask_model = MagnitudeMaskModel("ul_unas_16k")
wave_enhanced = mask_model.enhance(wave_noisy_16k)

Metrics

SeLib includes a few quick NumPy-only metrics that do not require PESQ/STOI or other external scoring libraries.

from selib.metrics import snr, segsnr, si_sdr

# Compare clean reference speech against an enhanced waveform.
print("SNR:", snr(wave_clean, wave_enhanced))
print("segSNR:", segsnr(wave_clean, wave_enhanced, sample_rate=sr))
print("SI-SDR:", si_sdr(wave_clean, wave_enhanced))

Available Models

Model ID	Type	Sample rate	#params	Paper	Repository
deepfilternet3	DeepFilterNet	48 kHz	2.13M	arXiv:2305.08227	GitHub
deepfilternet2	DeepFilterNet	48 kHz	2.31M	arXiv:2205.05474	GitHub
deepfilternet1	DeepFilterNet	48 kHz	1.78M	arXiv:2110.05588	GitHub
ul_unas_16k	Magnitude mask	16 kHz	0.171M	arXiv:2503.00340	GitHub

DeepFilterNet models use ERB masks plus deep-filter coefficients. Magnitude-mask models predict a spectrogram mask and reconstruct audio with the noisy phase.