Andrea Ramazzina

Papers

Total Citations

H-Index

About

Andrea Ramazzina is an emerging researcher working at the intersection of neural scene representation, sensor fusion, and robust perception for autonomous systems. Their work focuses on enabling reliable 3D understanding in challenging real-world conditions — including adverse weather and diverse sensor modalities — making their contributions highly relevant to autonomous vehicles, drones, and robotics. Ramazzina's most notable contribution, "ScatterNeRF" (2023, 25 citations), presents a physically-based inverse neural rendering framework that explicitly models light scattering and attenuation caused by fog, rain, and snow. By grounding the approach in physics, ScatterNeRF achieves robust scene reconstruction where conventional methods fail — a critical capability for safety-sensitive autonomous systems. Building on this momentum, their 2024 work "Radar Fields" (11 citations) extends neural field representations to FMCW radar, a sensing modality known for its resilience to weather but historically underexplored in neural reconstruction pipelines. This work fills a meaningful gap, enabling novel radar view synthesis and scene understanding beyond traditional RGB and LiDAR methods. Together, these contributions position Ramazzina as a thoughtful researcher pushing the boundaries of perception under uncertainty, with growing influence in the autonomous systems and computer vision communities.