Natalia Neverova

Meta (Israel), Centre National de la Recherche Scientifique

Papers

Total Citations

279

H-Index

About

Natalia Neverova is a computer vision researcher whose work sits at the intersection of scene understanding, temporal prediction, and human-robot interaction. She is perhaps best known for her influential 2017 paper "Predicting Deeper into the Future of Semantic Segmentation," which has accumulated over 235 citations and represents a significant advance in anticipatory visual perception. This work addresses a fundamental challenge in autonomous systems — the ability to forecast how a scene's semantic layout will evolve over time — making it directly relevant to applications in autonomous driving and robotics, where milliseconds of anticipatory understanding can be critical for safe decision-making. Rather than simply predicting raw pixel values, Neverova's approach operates at the level of semantic meaning, enabling richer and more actionable scene forecasts. Earlier in her career, she contributed to human-robot interaction research through the GEE corpus project, exploring naturalistic gesture recognition for socially intelligent robots. Across her body of work, Neverova demonstrates a consistent interest in bridging perception and action — equipping machines not just to see the world, but to understand and anticipate it in ways that support real-world deployment.