Research

I enjoy thinking about how subsystems and algorithmic approaches interact, and how these interactions can drive more sophisticated and robust autonomous systems. My research interests include object-level estimation, SLAM, deep learning, and autonomous navigation.

Katherine Liu 3/11/19 Katherine Liu 3/11/19

Weakly Supervised Deep Object Estimation

VoluMon is a novel approach to enable a deep neural network to predict the size and pose of objects without requiring 3D annotations or mesh models.

[pdf, video]

Katherine Liu 3/11/19 Katherine Liu 3/11/19

Object-Level SLAM for Autonomous Navigation

Robust Object-Based SLAM for High-Speed Autonomous Navigation (ROSHAN) improves object-level SLAM with ellipsoid landmarks for autonomous navigation by incorporating texture and semantic measurements.

[pdf]

Katherine Liu 3/11/19 Katherine Liu 3/11/19

Planning in Hybrid Geometric and Object-Level Representation

We enable more efficient navigation by training a neural network to predict a sampling distribution from object-level maps and occupancy maps.

[pdf][video]

Katherine Liu 3/11/19 Katherine Liu 3/11/19

Learning Gaussian Noise Models for State Estimation

Deep Inference for Covariance Estimation (DICE) uses a deep convolutional neural network to predict noise models for high dimensional, complex sensors.

[pdf][video]

Katherine Liu 7/3/18 Katherine Liu 7/3/18

Research

Weakly Supervised Deep Object Estimation

Object-Level SLAM for Autonomous Navigation

Planning in Hybrid Geometric and Object-Level Representation

Learning Gaussian Noise Models for State Estimation

Additional Publications