Papers
2
Total Citations
10
H-Index
2
About
Luca Crupi is a rising researcher at the forefront of autonomous robotics and embedded artificial intelligence, with a focus on enabling intelligent perception aboard resource-constrained platforms. His primary research areas include deep neural network architecture search, visual pose estimation, and self-supervised learning for robot localization. In his highly cited 2023 work, Crupi pioneered a neural architecture search method for accurate visual pose estimation aboard nano-UAVs—palm-sized drones with sub-100-gram payloads—demonstrating that deep learning can run efficiently on minuscule, low-power hardware. This contribution, garnering 7 citations, addresses a critical bottleneck in miniaturized robotics: achieving robust perception with extremely limited computational resources. More recently, in 2024, Crupi introduced a novel self-supervised learning approach for visual robot localization that leverages LED state prediction as a pretext task. By requiring only a few labeled samples alongside cheaply collected LED-state data, this method significantly reduces the annotation burden while maintaining high localization accuracy. With 3 citations already, this work showcases his talent for designing elegant, data-efficient solutions. Crupi’s research is paving the way for safer, more autonomous nano-drones capable of operating in human environments, marking him as a promising young innovator in embodied AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2