Elia Cereda

University of Applied Sciences and Arts of Southern Switzerland

Papers

Total Citations

H-Index

About

Elia Cereda is an emerging researcher at the forefront of embedded artificial intelligence, autonomous robotics, and tiny machine learning (TinyML), with a particular focus on vision-based perception for resource-constrained systems. His work tackles one of the most demanding challenges in modern robotics: enabling sophisticated deep learning capabilities on ultra-lightweight platforms, such as nano-drones weighing under 50 grams and operating within strict power budgets of just 20 milliwatts. Cereda's most cited contribution, "Vision-state Fusion" (2024, 11 citations), advances end-to-end deep neural network architectures by integrating visual and state information, improving perception for high-stakes applications including acrobatic UAV maneuvers and robot-assisted surgery. Complementing this, his research on on-device self-supervised learning aboard nano-quadrotors addresses a critical real-world limitation — the degradation of deployed models in previously unseen environments — by enabling drones to adapt autonomously without cloud connectivity or external supervision. Across his published work, Cereda demonstrates a consistent drive to push the boundaries of what miniaturized cyber-physical systems can achieve intelligently and independently. Though early in his career, his contributions are already shaping the trajectory of autonomous micro-robotics and efficient edge AI.