6 papers
Geometrically Plausible Object Pose Refinement using Differentiable Simulation
Anil Zeybek, Rhys Newbury, Snehal Dikhale +3
State-of-the-art object pose estimation methods are prone to generating geometrically infeasible pose hypotheses. This problem is prevalent in dexterous manipulation, where estimat…
KeyPointDiffuser: Unsupervised 3D Keypoint Learning via Latent Diffusion Models
Rhys Newbury, Juyan Zhang, Tin Tran +2
Understanding and representing the structure of 3D objects in an unsupervised manner remains a core challenge in computer vision and graphics. Most existing unsupervised keypoint m…
Heteroscedasticity of Denoising Score Matching with Generalised Smooth Noise
Juyan Zhang, Rhys Newbury, Xinyang Zhang +3
Score Matching (SM) is a powerful framework for estimating the log-density derivatives of a distribution without calculating its normalizing constants. This capability has made it…
INSPIRE-GNN: Intelligent Sensor Placement to Improve Sparse Bicycling Network Prediction via Reinforcement Learning Boosted Graph Neural Networks
Mohit Gupta, Debjit Bhowmick, Rhys Newbury +3
Accurate link-level bicycling volume estimation is essential for sustainable urban transportation planning. However, many cities face significant challenges of high data sparsity d…
Carefully Structured Compression: Efficiently Managing StarCraft II Data
Bryce Ferenczi, Rhys Newbury, Michael Burke +1
Creation and storage of datasets are often overlooked input costs in machine learning, as many datasets are simple image label pairs or plain text. However, datasets with more comp…
A Review of Differentiable Simulators
Rhys Newbury, Jack Collins, Kerry He +4
Differentiable simulators continue to push the state of the art across a range of domains including computational physics, robotics, and machine learning. Their main value is the a…