News

[21-Dec-2022] Starting Jan 2023 I will be an Associate Editor for the International Journal of Robotics Research (IJRR) in the area of “Localization and Mapping”. I look forward to reviewing some of the very best work being done in robotics :-)!

[13-Dec-2022] My research group has a new website! Check it out for all the latest and greatest news on trustworthy autonomy :-).

[7-Dec-2022] My research group has been awarded a new grant from Toyota Research Institute to develop novel perception and learning methods for high-agility autonomous driving!

[18-Oct-2022] Our paper on fast verification for certifiable estimation has been accepted to the IEEE Robotics and Automation Letters!

[8-Oct-2022] I delivered a keynote on our group’s work on certifiable machine perception in the workshop on “Safe and Certifiable Navigation and Control for Intelligent Vehicles in Complex Urban Scenarios” at this year’s IEEE International Conference on Intelligent Transportation Systems (ITSC). I very much enjoyed the lively discussion, as well as the opportunity to hear all about the great work being done in this area. Thanks to the organizers for assembling a fantastic series of talks (and also mine ;-P), and in particular to Dr. Weisong Wen for the invitation to speak.

[6-Oct-2022] We presented some of our group’s work at Northeastern’s Showcase of Opportunities for Undergraduate Research and Creative Endeavor SOURCE conference. Had a great time discussing math, computation, and robots with some of NEU’s aspiring young scholars :-)!

[14-Sep-2022] Our new paper presents the first convex relaxation for the general multi-robot range-aided SLAM (RA-SLAM) problem. We show how to reformulate RA-SLAM as a quadratically-constrained quadratic program (QCQP), which can then be relaxed to a second-order cone program (SOCP). Experimental evaluation demonstrates that this convex relaxation strategy enables the recovery of high-quality solutions on very challenging single- and multi-robot RA-SLAM problems involving thousands of poses and range measurements. Congrats to lead author Alan Papalia for this very cool piece of work :-)!

[14-Sep-2022] Our new paper describes a computationally-tractable method for approximating the full Bayesian posterior distribution in challenging high-dimensional, nonlinear, and non-Gaussian inference problems. Congrats to lead author Chad Huang!

[13-Sep-2022] I delivered the first installment of this year’s Research Seminar in Mathematics at Northeastern University, and had a great time talking math + robots / AI with our aspiring young mathematicians :-). Many thanks to Jonathan Weitsman for the invitation to speak!

[24-Aug-2022] I delivered a short talk on synergies between probability, optimization, and estimation & machine learning as part of this year’s annual CLEVR-AI MURI meeting. Many thanks to Mario Sznaier for the invitation to speak at this fantastic workshop :-)!

[18-Jul-2022] The SE-Sync library is now faster and more robust, thanks to the computational enhancements described in our latest paper. This latest update also includes some new visualizations of the optimization process that I think are pretty neat :-). Check it out!

[12-Jul-2022] Our new paper shows how to enhance the numerical speed and robustness of certifiable estimation methods using preconditioned eigensolvers.**

[30-Jun-2022] Two papers accepted to IROS 2022! Our work on spectral sparsification applies spectral graph theory to devise a principled (and very fast!) approach to pose-graph compression while preserving the accuracy of estimated robot poses, while our work on distributed Riemannian optimization provides a scalable communication-efficient approach to collaborative geometric estimation. Congrats to lead authors Kevin Doherty and Yulun Tian!

[27-May-2022] Had a great time at ICRA 2022! Enjoyed reconnecting with friends and colleagues, saw lots of great talks, presented two papers (on performance guarantees for spectral initialization in SLAM & rotation averaging and distance-geometric methods for inverse kinematics, respectively) and a preliminary abstract (on combining neural implicit surface representations with nonparametric Bayesian inference methods), and received an Honorable Mention for the King-Sun Fu Memorial Best Paper Award for our work on distributed certifiably correct pose-graph optimization (DC2-PGO). Looking forward to next year in London :-)!

[20-Apr-2022] Our paper on distributed certifiably correct SLAM has been recognized with an Honorable Mention for the IEEE Transactions on Robotics’s King-Sun Fu Memorial Best Paper Award! Thanks again to fantastic collaborators Kasra Khosoussi, Jonathan How, and especially lead author Yulun Tian for all of their hard work on this project, and to the T-RO editorial staff for their kind consideration :-).

[13-Apr-2022] I delivered a talk on certifiably correct machine perception at Worcester Polytechnic Institute’s Data Science Colloquium. Thanks to Oren Mangoubi for the invitation!

[12-Apr-2022] I delivered a talk on certifiably correct machine perception at Northeastern University’s Applied and Interdisciplinary Mathematics (AIMS) Seminar. Thanks to Christopher King for the invitation!

[25-Mar-2022] Our new paper applies a combination of spectral graph theory and convex relaxation to derive a simple, efficient, and information-theoretically sound method for compressing pose-graph representations of the SLAM problem. Congrats to lead author Kevin Doherty for a very cool piece of work :-)!

[2-Mar-2022] Our new paper presents the first distributed optimization algorithm with lazy communication for solving large-scale collaborative geometric estimation problems on arbitrary Riemannian manifolds. Congrats to lead author Yulun Tian for a very nice piece of work :-)!

[2-Feb-2022] I am an Associate Editor for IROS 2022. Looking forward to seeing some outstanding presentations in Kyoto this year – send us your very best work!

[31-Jan-2022] Our new paper on performance guarantees for spectral relaxations of pose-graph SLAM and rotation averaging will appear at ICRA 2022! Congrats to lead author Kevin Doherty for a very cool piece of work :-)

[28-Jan-2022] The SE-Sync library is now even easier to use thanks to some new Python bindings! Check out this Jupyter notebook for a minimal working example.

[5-Jan-2022] Our new RA-L paper describes how challenging inverse kinematics problems for highly-redundant manipulators can be efficiently and reliably solved through a clever combination of problem reformulation and convex relaxation. Kudos to lead authors Matt Giamou and Filip Maric for a very nice piece of work :-)!

[10-Dec-2021] I delivered a talk on certifiably correct machine perception at Hong Kong Polytechnic University’s Department of Aeronautical and Aviation Engineering (AAE) Seminar Series. Thanks to Weisong Wen for the invitation!

[23-Aug-2021] I am now an Assistant Professor at Northeastern University, with a joint appointment in the Departments of Electrical & Computer Engineering (ECE) and Mathematics. Looking forward to building lots of cool robots with my outstanding new colleagues :-)!

[05-May-2021] I delivered a talk on certifiably correct machine perception at the Technion Autonomous Systems Program (TASP) seminar series. Thanks to Vadim Indelman for the invitation!

[06-Jan-2021] Our new review of recent advances in SLAM is now available online! Many thanks to Kevin Doherty, Antonio Teran, and John Leonard for all of their hard work putting this together!

[05-Jan-2021] Our new paper presenting the first distributed certifiably-correct algorithms for SLAM and rotation averaging has been published in the IEEE Transactions on Robotics. Congrats to lead author Yulun Tian for a beautiful piece of work :-)!

[07-Dec-2020] I delivered a talk on certifiably correct SLAM at the University of Toronto’s Robotics Institute Seminar Series. Thanks to Jonathan Kelly for the invitation!

[25-Aug-2020] Our new paper presents the first distributed certifiably correct algorithms for pose-graph SLAM and rotation averaging

[16-Jul-2020] Our paper on learning to estimate rotations has received the Best Student Paper Award at Robotics: Science and Systems (RSS) 2020! Congrats to lead authors Valentin Peretroukhin and Matt Giamou for a very cool piece of work :-)!

[02-Jul-2020] Our paper on certifiably correct rotation averaging has been accepted as a spotlight talk (top 5%) at the European Conference on Computer Vision (ECCV) 2020!

[05-May-2020] New paper on learning to estimate rotations to appear at Robotics: Science and Systems (RSS) 2020

[08-Apr-2020] New paper on large-scale low-rank semidefinite programming to appear at the International Workshop on the Algorithmic Foundations of Robotics (WAFR) 2020

[02-Feb-2020] New paper on large-scale background modeling with a moving camera to appear at the Conference on Computer Vision and Pattern Recognition (CVPR) 2020

[23-Apr-2019] I have been selected as an RSS Pioneer!

[28-Feb-2019] New journal article on the SE-Sync algorithm for certifiably correct SLAM has been published in the International Journal of Robotics Research’s special issue!

[17-Sep-2018] I am now a postdoc in the Laboratory for Information and Decision Systems (LIDS) at the Massachusetts Institute of Technology. Looking forward to doing lots of cool math with my outstanding new colleagues :-)!