Publications
Real2Sem2Real: 3D Semantic Augmentations for Robot Learning from a Single Human RGB Demonstration
In submission
Turns one human RGB video into thousands of robot demonstrations across a whole object category.
VisTacAlign: Co-Training Dexterous Policies on Tactile Human and Robot Demonstrations
arXiv 2026
Co-trains dexterous policies on human and robot demonstrations by aligning vision, hand motion and touch.
J-PARSE: Jacobian-based Projection Algorithm for Resolving Singularities Effectively in Inverse Kinematic Control of Serial Manipulators
T-RO 2026
Smooth inverse kinematics near singularities, so manipulators can use more of their workspace in servoing, teleoperation and learning.
UniQueR: Unified Query-based Feedforward 3D Reconstruction
ECCV 2026
Feedforward 3D reconstruction as sparse 3D queries, inferring occluded geometry in a single forward pass.
MR. CRABS: Mobile RGB-D Camera-LIDAR Robot for Autonomous Bimanual Manipulation & Sensing
ICRA 2026 Workshop
A bimanual mobile robot with depth cameras, LIDAR and a pan-tilt camera for near-ground manipulation and search.
Learning to Drive is a Free Gift: Large-Scale Label-Free Autonomy Pretraining from Unposed In-The-Wild Videos
CVPR 2026
Pretrains a driving model on unlabeled, unposed videos, reaching state-of-the-art planning from a single front camera.
TensorTouch: Calibration of Tactile Sensors for High Resolution Stress Tensor and Deformation for Dexterous Manipulation
T-RO 2026
Calibrates optical tactile sensors to estimate stress tensors, deformation and force at pixel resolution.
SPACeR: Self-Play Anchoring with Centralized Reference Models
ICLR 2026
Guides self-play driving RL with a pretrained motion model, matching human driving with 10x faster inference.
DexFruit: Dexterous Manipulation and Gaussian Splatting Inspection of Fruit
RA-L 2025
Gentle robotic handling of fragile fruit with optical tactile sensing, plus 3D damage inspection with Gaussian splats.
Next Best Sense: Guiding Vision and Touch with FisherRF for 3D Gaussian Splatting
ICRA 2025 Oral
Chooses where to look and touch next to train few-shot Gaussian splats, guided by FisherRF.
Touch-GS: Visual-Tactile Supervised 3D Gaussian Splatting
IROS 2024 Oral
Combines vision and touch to build Gaussian splats of hard scenes: few views, mirrors, transparent objects.
SHIRO: Soft Hierarchical Reinforcement Learning
arXiv 2022
Soft hierarchical reinforcement learning that speeds up learning on hard locomotion tasks.
Individualized Empirical Baselines for Evaluating the Energy Performance of Existing Buildings
STBE 2022
Empirical baselines for evaluating the energy performance of existing buildings.
Evaluating Performance of Different Generative Adversarial Networks for Large-Scale Building Power Demand Prediction
Energy and Buildings 2022
Compares types of GANs for large-scale building power demand prediction.
Volumetric Data Fusion of External Depth and Onboard Proximity Data For Occluded Space Reduction
IROS 2021 Workshop
Fuses external depth and onboard proximity data into a 3D map, shrinking the space a robot must treat as unknown.
Self-Contained Kinematic Calibration of a Novel Whole-Body Artificial Skin for Human-Robot Collaboration
IROS 2021
Self-calibration of a robotic skin placed anywhere on an arm, for physical human-robot interaction.
Contact Anticipation for Physical Human-Robot Interaction with Robotic Manipulators using Onboard Proximity Sensors
IROS 2021
Onboard proximity sensors let a robot arm anticipate and react to contact with people.
Development of New Baseline Models for U.S. Medium Office Buildings Based on Commercial Buildings Energy Consumption Survey Data
STBE 2020
New energy-use baseline models for U.S. medium office buildings in different environments.
Theses
Enabling Close Proximity Human Robot Collaboration via Distributed, Self-Calibrating Robotic Skin
Undergraduate Thesis 2021
My undergraduate thesis, covering two years of research on robotic skin for close human-robot collaboration.