Tensorflow 2.x implementation of the stacked dual-signal transformation LSTM network (DTLN) for real-time noise suppression. This repository provides the code for training, infering and serving the DTLN model in python. It also provides pretrained models in SavedModel, TF-lite and ONNX format,…
Unofficial PyTorch implementation of Google AI's VoiceFilter system
rnnoise is a real-time speech denoising library that uses a recurrent neural network to suppress background noise from live voice audio. It is implemented as a lightweight C library with a minimal API, designed for easy integration into audio applications that need low-latency noise reduction. The library employs a gated recurrent unit (GRU) architecture and frequency-domain feature extraction to capture temporal dependencies in speech, operating on short audio frames sequentially for streaming use. It also includes a training pipeline that allows users to train custom noise suppression model