Peer-reviewed work spanning generative-model representation learning and signal processing for wireless sensor networks. Each links to the published paper.
A method for disentangling generative latent representations using orthogonal latent codes and an inter-domain signal transformation, giving finer control and interpretability over the latent space. Published in the proceedings of ICAART 2023, with Prof. Emanuele Rodolà (Sapienza).
View paperAn adaptive environmental-modeling approach that improves RSSI-based node localization in wireless sensor networks without the extra hardware that range-based methods usually need. Published at IEEE ICEE 2012, from my MSc thesis (final grade 19.9/20).
View paperInterested in the research behind these? Get in touch.