Elisabetta Delponte

Papers

Total Citations

H-Index

About

Elisabetta Delponte’s research centers on computer vision, specifically appearance-based 3D object recognition and the integration of temporal information with sparse image representations. Her most cited work, a 2007 paper, explores how temporally dense view-based recognition can be combined with local keypoint features to extract time-invariant characteristics that distinguish objects across varying viewpoints. This contribution addresses a fundamental challenge in robotics and autonomous systems: enabling reliable object identification as an observer moves through space. By leveraging temporal continuity, Delponte’s approach enhances the robustness of recognition systems, allowing them to maintain accuracy even when objects are partially occluded or viewed from unfamiliar angles. Though her citation counts are modest (4 and 3 citations for her top papers), her work represents an early and thoughtful attempt to bridge the gap between dynamic visual input and static feature extraction—a problem that remains central to modern deep learning-based vision systems. Her research is particularly relevant for students and engineers working on real-time perception for drones, mobile robots, or augmented reality, where understanding how objects appear over time is as critical as recognizing them in a single frame.