Viviana Crescitelli

Papers

Total Citations

H-Index

About

Viviana Crescitelli is a researcher at the forefront of computer vision and human-robot interaction, with a primary focus on enabling collaborative robots to perceive and predict human motion in challenging, real-world environments. Her work addresses a critical gap in human pose estimation (HPE), where most state-of-the-art deep learning models rely on standard RGB images and fail under low-light conditions. In her most-cited paper, "An RGB/Infra-Red camera fusion approach for Multi-Person Pose Estimation in low light environments" (2020, 10 citations), Crescitelli pioneered a multi-modal sensor fusion technique that combines RGB and infrared imagery. This approach allows deep convolutional neural networks to robustly estimate anatomical keypoints even in darkness, significantly expanding the operational scope of collaborative robotics. By tackling the overlooked problem of illumination variance, her contributions enhance the safety and reliability of human-robot teams in industrial and service settings. Crescitelli’s work is notable for bridging the gap between theoretical computer vision and practical deployment, demonstrating a keen understanding of the real-world constraints that limit current AI systems. Her research continues to inspire new directions in sensor fusion and robust perception for autonomous systems.