We develop intelligent, reliable, and resource-efficient learning systems combining signal processing, artificial intelligence, and autonomous decision-making for real-world and industrial applications. Our research focuses on three core areas:

• Signal Processing for Localization and Sensing: We develop data-driven methods for localization, tracking, motion prediction, and multimodal sensor fusion, emphasizing self-supervised learning and integrated sensing and communication for 6G and beyond.
• Edge AI and Resource-Efficient Learning: We design accurate, memory- and energy-efficient learning algorithms for resource-constrained devices, including embedded systems and microcontrollers.
• Reliable and Trustworthy Reinforcement Learning: We develop safe, stable, and interpretable reinforcement learning methods for uncertain, industrial, and safety-critical environments.

You can find more about our research here.