Problem: Due to environmental changes such as temperature changes and multi-path fading effect, the wireless signals can significantly vary from time to time (as shown in the figure), causing the localization accuracy to drop.

Our Solution: We address this problem by introducing a transferred Hidden Markov Model (TrHMM).  In TrHMM, we aim to transfer out-of-date model to fit a current model through learning, even though the training data have very different distributions.

Dataset and Code: Download data. Download code. Download paper.
　

Problem: Traditional indoor localization systems often assume that the collected signal data distributions are fixed, ragardless of device hardware difference. However, by empirically studying the signal variation over different devices, we found this assumption to be invalid in practice (as shown in the figure).

Our Solution: We propose a latent multi-task learning (LatentMTL) algorithm, which treats multiple devices as multiple learning tasks. In LatentMTL, we require the hypotheses learned in a latent feature space are similar; and we employ alternating optimization to iteratively learn feature mappings and multi-task regression models.

Dataset and Code: Download data. Download code. Download paper.