Valentina Donzella

University of Warwick, Queen Mary University of London

Papers

Total Citations

H-Index

About

Valentina Donzella is a leading researcher at the intersection of autonomous systems, computer vision, and safety-critical AI. Her work focuses on ensuring that vision-based robots and vehicles can operate reliably in complex, dynamic environments. She has made major contributions to camera image simulation, a cornerstone for the virtual validation of autonomous vehicles. Her 2023 paper on accelerating stereo image simulation using neural super-resolution techniques addresses the critical need for high-fidelity, efficient synthetic data generation, which is essential for training robust perception models. Donzella is also pioneering approaches to safety assurance for AI-driven perception systems. Her 2025 work introduces a novel situation coverage methodology to systematically test the robustness of vision-based AI against real-world challenges like sensor degradation and environmental variability. While her most cited papers are recent, their impact is growing rapidly within the autonomous driving and robotics communities. Donzella’s research is vital for bridging the gap between simulation and reality, ultimately enabling safer deployment of autonomous technologies.