Maria Bauza Villalonga

I am PhD student at the Massachusetts Institute of Technology (MIT) working with Prof. Alberto Rodriguez.

I develop algorithms and solutions that enable robots to solve new tasks with high accuracy and dexterity. My research has been supported by LaCaixa and Facebook fellowships.  /  CV  /  Bio  /  Google Scholar  /  LinkedIn    

profile photo

I develop algorithms that enable embodied intelligence to make robots perceive and interact with their environment accurately. In my work, I have studied both the power of applying the most recent advances in AI and computer vision to control robotic systems and the capabilities of high-resolution tactile sensing to solve complex tasks, such as grasping, localization, and precise placing. My goal is to effectively leverage AI on any source of sensing to make robots more accurate, reactive, and dexterous.


[UPCOMING] December 11th, 2021 Invited talk at Washington University robotics colloquium.

[UPCOMING] November 14th, 2021 Invited talk at Stanford.

[UPCOMING] November 5h, 2021 Invited talk at CMU.

[UPCOMING] October 28th, 2021 Invited talk at Cornell's Robotic Seminar.

[UPCOMING] October 2021 Selected to attend the Rising Stars in EECS.

Sept 2021 Accepted paper Tailoring at 2021 NeurIPS.

July 2021 Attended the 2021 RSS Pioneers Workshop.

May 2021 Best Paper Finalist Award on Service Robotics at ICRA 2021

March 2021 Invited talk at University of Toronto, AI in Robotics Seminar.

October 2020 Invited talk at University of Pennsylvania, Grasp Seminar.

May 2020 Co-organizing workshop at ICRA 2020 on Uncertainty in Contact-Rich Interactions (canceled due to CoVID19).

October 2019 Selected to attend the Global Young Scientists Summit. Awarded to only 5 PhDs from all MIT departments.

October 2019 Rising Stars in Mechanical Engineering. Awarded to 30 graduate and postdoctoral women.

2018 Awarded the Facebook Emerging Scholar Award. 21 awardees out of more than 900 applications.

2018 Awarded the NVIDIA Graduate Fellowship. Given to 10 PhD students from more than 230 applications.

Tactile Object Pose Estimation from the First Touch with Geometric Contact Rendering
M. Bauza, E. Valls, B. Lim, T. Sechopoulos
CORL, 2020
PDF / video / website

We learn in simulation how to accurate localize objects. Our solution transfers to the real world, localizing objects with tactile from the first touch.

This technology is used by Magna and ABB. Our tactile sensor is Gelslim.

Tactile Mapping and Localization from High-Resolution Tactile Imprints
M. Bauza, O. Canal, A. Rodriguez
ICRA, 2019
PDF / video / website

Shape reconstruction and object localization using the vision-based tactile sensor GelSlim.

Real-time shape and pose estimation from planar pushing using implicit surfaces
S. Suresh, M.Bauza, A. Rodriguez, J. Mangelson, M. Kaess
ICRA, 2021   (Best Paper Finalist on the ICRA21 Service Robotics Award)
PDF / video / code / website

In real-time, we infer from planar pushes both the shape and pose of an object.

Tailoring: Encoding Inductive Biases by Optimizing Unsupervised Objectives at Prediction Time
F. Alet, K. Kawaguchi, M. Bauza, N. Kuru, T. Lozano-Perez, L. Kaelbling
NeurIPS, 2021

Accurate Vision-based Manipulation through Contact Reasoning
A. Kloss, M. Bauza, J. Wu, J. Tenenbaum, A. Rodriguez, J. Bohg
ICRA, 2020
PDF / video

Experience-Embedded Visual Foresight
Y. Lin, M. Bauza, P. Isola
CORL, 2019
PDF / code / website

Leaning to encode new objects to generate physically plausible video predictions.

Tactile Regrasp: Grasp Adjustments via Simulated Tactile Transformations
M. Bauza*, F. Hogan* , O. Canals, A. Rodriguez
IROS, 2018   (Best Poster Award at ICRA 2018 workshop)
PDF / video

Tactile regrasping using a high-resolution tactile sensor to improve grasp stability.

Graph Element Networks: adaptive, structured computation and memory
F. Alet, A. Jeewajee, M. Bauza, A. Rodriguez, T. Lozano-Perez, L. Kaelbling
ICML, 2019   (Oral Presentation)
PDF / video / website

Omnipush: accurate, diverse, real-world dataset of pushing dynamics with RGB-D video
M. Bauza, F. Alet, Y. Lin, T. Lozano-Perez, L. Kaelbling, P. Isola, A. Rodriguez
IROS, 2019
PDF / website

We present a large high-quality on planar pushing that includes RGB-D video and extense object variability.

Learning vs. physics-based control of a planar push system
M. Bauza*, F. Hogan* , A. Rodriguez
CORL, 2018
PDF / video

We explore the data-complexity required for controlling, rather than modeling, planar pushing.

Combining Physical Simulators and Object-Based Networks for Control
A. Ajay, M. Bauza, J. Wu, N. Fazeli, J. Tenenbaum, A. Rodriguez
ICRA, 2019
PDF / website

We propose a hybrid dynamics model, simulator-augmented interaction networks (SAIN), combining a physics engine with an object-based neural network for dynamics modeling.

Augmenting Simulators with Stochastic Networks
A. Ajay, J. Wu, N. Fazeli, M. Bauza, L. Kaelbling, J. Tenenbaum, A. Rodriguez
IROS, 2018   (Best Paper Award on Cognitive Robotics)
PDF / website

We augment an analytical rigid-body simulator with a neural network that learns to model uncertainty as residuals. Best Paper Award on Cognitive Robotics at IROS 2018.

Active Perception of Novel Objects in Clutter with Multi-Affordance Grasping and Cross-Domain Image Matching
A Zeng, S Song, K. Yu, E. Donlon, F. Hogan, M. Bauza, et. al.
ICRA, 2018   (Best Systems Paper Award by Amazon Robotics)
PDF / video / website

With the MIT-Princeton team we developed a robust robotic system for bin picking.

GP-SUM. Gaussian Processes Filtering of non-Gaussian Beliefs
M. Bauza, A. Rodriguez
WAFR, 2018

We developed the algorithm GP-SUM: a GP-Bayes filter that propagates in time non-Gaussian beliefs.

A Probabilistic Data-Driven Model for Planar Pushing
M. Bauza, A. Rodriguez
ICRA, 2017

Characterizing the uncertainty of different pushes allows better action selection.

More than a Million Ways to Be Pushed. A High-Fidelity Experimental Data Set of Planar Pushing
K. Yu, M. Bauza, N. Fazeli, and A. Rodriguez
IROS, 2016   (Best Paper Finalist at IROS)
PDF / video / website

More than a million datapoints collected on real pusing experiments.

Thanks to Jon Barron for sharing his web design.