Sudharshan Suresh

suddhu [at] cmu [dot] edu

I'm a final-year PhD candidate in the Robotics Institute at Carnegie Mellon University, advised by Michael Kaess. I work on spatial AI from vision and touch for robot manipulation. My thesis focuses on learning object-centric representations and harnessing vision-based touch. I'm also a part-time researcher at FAIR (Meta), where I work with Mustafa Mukadam.

I completed my Masters in Robotics at CMU with Michael Kaess on underwater exploration and SLAM (thesis). Prior to that, I worked with Red Whittaker on state-estimation for lunar rovers, and at IISc Bangalore working on visual understanding. In my undergrad, I majored in Controls and Instrumentation at NIT Trichy.

CV  /  Scholar  /  Github  /  LinkedIn  /  Twitter  /  Short bio


[Dec '23]    

The pre-print for NeuralFeels is out, read it here!

[Aug '23]    

Our work RotateIt, led by Haozhi, was accepted to CoRL 2023.

[April '23]    

Spending the summer as a research scientist intern at FAIR Menlo Park on visuo-tactile manipulation!

[Dec '22]  

MidasTouch was showcased at CoRL 2022 with a live demo.

[Oct '22]  

Successfully passed my Ph.D. thesis proposal!

[Sep '22]  

MidasTouch was accepted to CoRL 2022 as an oral.

[Aug '22]  

We've extended iSDF for neural mapping with the Franka robot, code here.

Click for more updates

[May '22]    

Organized the Debates on the Future of Robotics Research workshop at ICRA '22

[April '22]    

Spending the summer at FAIR Pittsburgh working on pose tracking from touch

[Jan '22]    

ShapeMap 3-D was accepted to ICRA 2022, with an open-source implementation.

[Aug '21]    

Presented at the Tartan SLAM series on our working on perception for planar pushing, video here.

[May '21]    

Tactile SLAM was the ICRA 2021 best paper in service robotics finalist!



Neural feels with neural fields: Visuo-tactile perception for in-hand manipulation
Sudharshan Suresh, Haozhi Qi, Tingfan Wu, Taosha Fan, Luis Pineda, Mike Lambeta, Jitendra Malik, Mrinal Kalakrishnan, Roberto Calandra, Michael Kaess, Joe Ortiz, and Mustafa Mukadam
Pre-print, Dec 2023
paper / website / presentation
Neural perception with vision and touch yields robust tracking
and reconstruction for in-hand manipulation
General In-Hand Object Rotation with Vision and Touch
Haozhi Qi, Brent Yi, Sudharshan Suresh, Mike Lambeta Yi Ma, Roberto Calandra, and Jitendra Malik
Proc. Conf. on Robot Learning, CoRL, Nov 2023
paper / website
A visuotactile transformer gives us general dexterity
for multi-axis object rotation in the wild.
MidasTouch: Monte-Carlo inference over distributions across sliding touch
[Oral: 6% acceptance rate]
Sudharshan Suresh, Zilin Si, Stuart Anderson, Michael Kaess, and Mustafa Mukadam
Proc. Conf. on Robot Learning, CoRL, Dec 2022
paper / website / code / presentation
Where's Waldo? but for robot touch: tracking a robot finger
on an object from geometry captured by touch.
ShapeMap 3-D: Efficient shape mapping through dense touch and vision
Sudharshan Suresh, Zilin Si, Joshua Mangelson, Wenzhen Yuan, and Michael Kaess
IEEE Intl. Conf. on Robotics and Automation, ICRA, May 2022
paper / website / code / presentation
Online reconstruction of 3D objects from dense touch
and vision via Gaussian processes.
Tactile SLAM: Real-time inference of shape and pose from planar pushing
[ICRA best paper award in service robotics finalist]
Sudharshan Suresh, Maria Bauza, Peter Yu, Joshua Mangelson, Alberto Rodriguez, and Michael Kaess
IEEE Intl. Conf. on Robotics and Automation, ICRA, May 2021
paper / website / presentation
Full SLAM from force/torque sensing for planar pushing:
combining a factor graph with an implicit surface.
Active SLAM using 3D submap saliency for underwater volumetric exploration
Sudharshan Suresh, Paloma Sodhi, Joshua Mangelson, David Wettergreen, and Michael Kaess
IEEE Intl. Conf. on Robotics and Automation, ICRA, May 2020
paper / presentation
Balancing volumetric exploration and pose uncertainty
in 3D underwater SLAM via SONAR submap saliency.
ARAS: ambiguity-aware robust active SLAM using multi-hypothesis estimates
Ming Hsiao, Joshua Mangelson, Sudharshan Suresh, Christian Debrunner, and Michael Kaess
IEEE Intl. Conf. on Intelligent Robots and Systems, IROS, Oct 2020
Active SLAM with multi-hypothesis state estimates
for robust indoor mapping with handheld sensors
Through-water stereo SLAM with refraction correction for AUV localization
Sudharshan Suresh, Eric Westman, and Michael Kaess
IEEE Robotics and Automation Letters (RA-L), presented at ICRA 2019, Jan 2019
paper / presentation
Dealing with refraction in underwater visual SLAM,
inspired by multimedia photogrammetry.
Localized imaging and mapping for underwater fuel storage basins
Jerry Hsiung, Andrew Tallaksen, Lawrence Papincak, Sudharshan Suresh, Heather Jones, Red Whittaker, and Michael Kaess
Proceedings of the Symposium on Waste Management, Phoenix, Arizona, Mar 2018
paper / slides / video
We build an underwater platform comprising of stereo,
IMU, standard + structured lighting, and depth.
Camera-Only Kinematics for Small Lunar Rovers
Sudharshan Suresh , Eugene Fang, and Red Whittaker
Robotics Institute Summer Scholars Working Paper Journal, Nov 2016
Annual Meeting of the Lunar Exploration Analysis Group, Nov 2016
paper / video / poster
Tracking a lunar rover's kinematic state through self-perception
with a downward-facing fisheye lens.
Object category understanding via eye fixations on freehand sketches
Ravi Kiran Sarvadevabhatla, Sudharshan Suresh and R. Venkatesh Babu
IEEE Transactions on Image Processing (TIP), May 2017
paper / website / dataset
We understand free-hand sketches through human gaze
fixations based on visual saliency.


Other projects

Franka iSDF: neural mapping for tabletop manipulation
Sudharshan Suresh, Joe Ortiz, and Mustafa Mukadam
Extending iSDF to build real-time neural models
of tabletop scenes with the Franka Panda arm
DeepGeo: photo localization with deep neural network
Sudharshan Suresh, Nate Chodosh, and Montiel Abello
arXiv / github
A deep network that beats humans at GeoGuessr,
trained on our 50States10K dataset
Task and motion planning for robotic food preparation
Sudharshan Suresh, Travers Rhodes, Montiel Abello, and Himanshi Yadav
pdf / video 1 / video 2
Hierarchical task and motion planning for a 6-DOF robot arm,
to prepare yogurt parfaits!
Thin structure reconstruction via 3D lines and points
Sudharshan Suresh and Montiel Abello
Reconstructing thin objects in a scene through an
SfM pipeline can be hard!
Factor graph optimization for dynamic parameter estimation
Sudharshan Suresh, Eric Dexheimer, and Montiel Abello
A method to estimate MAV poses and dynamic parameters
during flight.

Last updated: Nov 2023

Imitation is the highest form of flattery