Skip to content

aume/plcTorch

Repository files navigation

PLCModel.py

Dataset Preparation:

AudioDataset simulates packet loss by randomly masking parts of the audio signal.

Model Definition:

PLCModel is an RNN model that predicts missing audio packets.

Training Loop:

The model is trained to minimize the mean squared error between the predicted audio and the original audio where packets are present.

Real-Time Concealment:

The real-time script captures audio input, simulates packet loss, and uses the trained model to conceal the lost packets, playing back the reconstructed audio.

Ensure you have the necessary libraries (torch, torchaudio, sounddevice, sklearn) installed.

PLCModelLC.py

This script captures real-time audio input, simulates packet loss, and uses the trained model to reconstruct and output the audio, achieving real-time packet loss concealment.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages