speech enhancement mask resnet unet

Providerhuseinzol05
Categoryspeech-enhancement
LicenseApache-2.0
Downloads0
Stars0

Overview

This model implements a hybrid ResNet-U-Net architecture designed for speech enhancement via mask estimation. Rather than predicting clean audio directly, it learns to generate a time-frequency mask that filters out stationary and non-stationary noise from the input spectrogram. For developers, this means a robust approach to denoising that leverages the multi-scale feature extraction of U-Net with the gradient stability of residual connections. It is particularly suited for preprocessing pipelines in ASR (Automatic Speech Recognition) or VoIP applications where improving the Signal-to-Noise Ratio (SNR) is critical. Integration typically involves a STFT/ISTFT wrapper to handle the conversion between waveforms and the spectrograms the model processes.

Highlights

  • Hybrid ResNet-U-Net architecture for effective noise suppression
  • Uses mask-based estimation to isolate clean speech signals
  • Ideal for ASR preprocessing and real-time audio cleaning
  • Permissive Apache-2.0 license for commercial integration
  • Optimized for time-frequency domain speech enhancement

Usage

Install
# Install Hugging Face transformers
pip install transformers torch
SDK Usage
# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("huseinzol05/speech-enhancement-mask-resnet-unet")
tokenizer = AutoTokenizer.from_pretrained("huseinzol05/speech-enhancement-mask-resnet-unet")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance 
pip install -U huggingface_hub

CLI Download

Download the full repository

Download the full repository 
huggingface-cli download huseinzol05/speech-enhancement-mask-resnet-unet

Download a single file to a local folder (e.g. config.json into ./dir)

Download a single file to a local folder (e.g. config.json into ./dir) 
huggingface-cli download huseinzol05/speech-enhancement-mask-resnet-unet config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

SDK Download 
# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('huseinzol05/speech-enhancement-mask-resnet-unet')

Git Download

Make sure git-lfs is installed first

Git Download 
git lfs install
git clone https://huggingface.co/huseinzol05/speech-enhancement-mask-resnet-unet

To skip LFS large-file downloads, use:

Skip LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/huseinzol05/speech-enhancement-mask-resnet-unet

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

Install Transformers 
pip install -U transformers torch

Load the model and run inference

Load the model and run inference 
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('huseinzol05/speech-enhancement-mask-resnet-unet')
tokenizer = AutoTokenizer.from_pretrained('huseinzol05/speech-enhancement-mask-resnet-unet')
View official source
Join our Telegram