DeepFilterNet is a full-band audio low-complexity speech enhancement framework.

This framework supports Linux, macOS and Windows, and the structure of the framework is as follows:

• libDF Contains Rust code for data loading and augmentation
• DeepFilterNet Contains DeepFilterNet code training, evaluation and visualization, and pretrained model weights
• pyDF-data Contains a Python wrapper for libDF dataset functionality and provides a PyTorch data loader.
• ladspa A LADSPA plugin is included for real-time noise suppression.
• models Contains pre-training used in DeepFilterNet (Python) or libDF/deep-filter (Rust).

Instructions

deep-filter

Download precompiled binaries from the Release page, you can use deep-filter to suppress noise in noisy .wav audio files. Currently, only wav files with a sampling rate of 48kHz are supported.

USAGE:
    deep-filter [OPTIONS] [FILES]...

ARGS:
    <FILES>...

OPTIONS:
    -D, --compensate-delay
            Compensate delay of STFT and model lookahead
    -h, --help
            Print help information
    -m, --model <MODEL>
            Path to model tar.gz. Defaults to DeepFilterNet2.
    -o, --out-dir <OUT_DIR>
            [default: out]
    --pf
            Enable postfilter
    -v, --verbose
            Logging verbosity
    -V, --version
            Print version information