Semi-Supervised Learning with VAEs for Device-Free Fingerprinting Indoor Localization

Introduction:

Implementation of semi-supervised variational auto-encoder (VAE) for Device Free Wi-Fi Fingerprinting Indoor Localization.

For more details and evaluation results, please check out our original paper, which will be published in IEEE GLOBECOM 2021.

The code is inherited and modified from here.

main.py: train the VAE model under the pre-defined indoor localization scenarios.
plot_CSI.py: code for plotting CSI samples
models.py: definde semisupervised VAE
dataset: pre-collected CSI samples save as pickle in the form of (X_train, y_train, X_tst, y_tst)