Ivano Donadi
Papers
2
Total Citations
12
H-Index
2
About
Ivano Donadi is a computer vision researcher whose work bridges the gap between geometric reasoning and deep learning for robotics and autonomous systems. His primary research areas include object pose estimation, 3D scene understanding, and traversability analysis for autonomous navigation. Donadi’s most notable contribution is the development of **KVN (Keypoints Voting Network with Differentiable RANSAC)**, a novel framework for stereo pose estimation that integrates differentiable RANSAC into an end-to-end learning pipeline. This work, published in 2024 with 7 citations, addresses a fundamental challenge in robotics and augmented reality by enabling robust 2D-3D keypoint correspondence prediction and PnP-based pose estimation. In parallel, his 2023 work on **Pyramidal 3D Feature Fusion on Polar Grids** (5 citations) introduces a real-time, CPU-efficient method for traversability analysis that combines geometric features with machine learning, directly supporting safe navigation for self-driving vehicles and ground robots. Donadi’s research is characterized by its practical focus on real-time performance and hardware efficiency, making his methods directly applicable to deployed robotic systems. His work represents a meaningful step toward more reliable and computationally accessible perception for autonomous agents.
Research Focus
Key Achievements
Top Papers
- 1
- 2