Semidefinite relaxations, particularly the moment-SOS hierarchy (a.k.a. Lasserre's hierarchy), are a powerful tool for the global optimization of polynomial optimization problems. However, it is challenging to apply them to practical problems because they often lead to SDPs that are too large to...
Certifiable Outlier-Robust Geometric Perception: Robots that See through the Clutter with Confidence
Geometric perception is the task of estimating geometric models from sensor measurements and priors. The ubiquitous existence of outliers —measurements that...
I was featured by MIT Spotlight (front page of MIT) for "making self-driving cars safer through keener robot perception"!
See the spotlight here:
Here is the link to the talk title/abstract:
This is "the" paper to read if you want to know more about certifiable geometric perception and to design trustworthy algorithms for robot perception.
The paper "Optimal Pose and Shape Estimation for Category-level 3D Object Perception", joint work with Jingnan Shi and Luca Carlone, is a Best Paper Award Finalist at Robotics: Science and Systems (RSS) 2021.
Check out the paper:
I am happy to be selected as one of the 30 Pioneers at Robotics: Science and Systems (RSS) 2021, for my research on "certifiable outlier-robust machine perception".
Check out the full list of Pioneers here:
The paper "Graduated Non-Convexity for Robust Spatial Perception: From Non-Minimal Solvers to Global Outlier Rejection" has won an IEEE Robotics and Automation Letters Best Paper Award Honorable Mention. Joint work with Pasquale Antonante, Vasileios Tzoumas, and Luca Carlone.
I gave an invited talk at Imperial College London MatchLab on "Self-supervised Geometric Perception".
Slides are available here.
- 1 of 4
- next ›