Matteo Poggi
Papers
7
Total Citations
50
H-Index
4
About
Matteo Poggi is a computer vision researcher whose work centers on depth estimation, 3D reconstruction, and scene understanding — areas with profound implications for robotics, autonomous driving, and augmented reality. He has made significant contributions to unsupervised monocular depth estimation, notably advancing techniques capable of running in real time on CPU-constrained hardware, democratizing access to depth perception without reliance on expensive sensors. His research spans a broad methodological spectrum, from contrastive learning approaches to depth prediction and LiDAR confidence estimation, to dense SLAM systems enhanced with hybrid neural representations. Poggi has also tackled specialized challenges such as reconstructing tiny objects in industrial robotics settings and streaming dense depth from low-frame-rate sensors — pushing the boundaries of what is achievable with minimal hardware assumptions. His 2018 work on real-time unsupervised monocular depth estimation has garnered 19 citations, reflecting its foundational influence, while his more recent contributions continue to shape how the community approaches structured scene understanding. Across his portfolio, Poggi consistently bridges theoretical innovation with practical deployability, making his research particularly valuable for engineers and scientists working at the intersection of perception and real-world autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1Towards Real-Time Unsupervised Monocular Depth Estimation on CPU19 citations · 2018
- 2Contrastive Learning for Depth Prediction12 citations · 2023
- 3Unsupervised confidence for LiDAR depth maps and applications7 citations · 2022
- 4
- 5
- 6
- 7Real-Time Semantic Stereo Matching2 citations · 2020