Michael Kaess retweeted
We are excited to announce the release of our open-source calibration package called TartanCalib, which improves reprojection errors by up to 27%, helps us detect up to 42% more features, and improves performance in dense depth mapping: youtu.be/8WCCS-Ag_Co
A different approach to learning that doesn't require differentiable optimizers!
We will be talking about our work, LEO, this Wednesday at #CoRL2021. We show that learning observation models can be viewed as shaping energy functions that graph optimizers, even non-differentiable ones, optimize. Come visit us at the poster session!
Michael Kaess retweeted
Tomorrow at 12:00 pm EST @jleonardmit will be presenting his talk titled, "The Past, Present and Future of SLAM" live on YouTube (youtu.be/FH6suW6_A5U), followed by an open discussion.
You can check out the previous and future talks on our website: theairlab.org/tartanslamseri…
Congratulation to my students: 7 papers at ICRA including one award finalist!
Here's a look at our lab's research, presented at #ICRA2021. We have papers spanning object SLAM, monocular perception, manipulation, and optimization!
(1/7)
A Graph-Based Method for Joint Instance Segmentation of Point Clouds and Image Sequences
Paper: cs.cmu.edu/~kaess/pub/Abello…
Congratulations to all my students! More details on the papers soon.
Pleased to announce that all 7 papers submitted by RPL at #icra2021 have been accepted!
Be sure to also checkout the workshop on Visual Inertial Navigation systems:
udel.edu/~ghuang/icra21-vins…
Congratulations to all the authors!
Congratulations Ed!
Dear @NASAPersevere, what an amazing landing! Congratulations! My joy trebled when I saw that you have a number of AprilTag visual fiducials on board (github.com/AprilRobotics/apr…). I pulled out the AprilTag iOS app, enabled the 16h5 tag family, and was able to read these tags!
Michael Kaess retweeted
Here's a thread of the lab's research, presented at #IROS2020:
(1/5)
Efficient Multiresolution Scrolling Grid for Stereo Vision-based MAV Obstacle Avoidance
#iros2020 on-demand: iros2020.org/ondemand/episod…
Paper: cs.cmu.edu/~kaess/pub/Dexhei…
Michael Kaess retweeted
40 papers from the RI accepted to IROS 2020! Super proud of our researchers. Well done everyone! @CMU_Robotics
Michael Kaess retweeted
Congratulations to Michael Kaess @CMU_Robotics, who shares the inaugural @RoboticsSciSys Test of Time Award with Frank Dellaert of @GeorgiaTech, a Ph.D. alum of SCS. bit.ly/3eWJtYX
ALT Michael Kaess, associate research professor, Carnegie Mellon University Robotics Institute