Robert Platt

Latest

• Equivariant Diffusion Policy
• ThinkGrasp: A Vision-Language System for Strategic Part Grasping in Clutter
• IMAGINATION POLICY: Using Generative Point Cloud Models for Learning Manipulation Policies
• Fourier Transporter: Bi-Equivariant Robotic Manipulation in 3D
• Leveraging Pick and Place Symmetries
• A General Theory of Correct, Incorrect, and Extrinsic Equivariance
• One-shot Imitation Learning via Interaction Warping
• Equivariant Reinforcement Learning under Partial Observability
• On Robot Grasp Learning Using Equivariant Models
• The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetry
• Image to Sphere: Learning Equivariant Features for Efficient Pose Prediction
• SEIL: Simulation-augmented Equivariant Imitation Learning
• On-Robot Learning With Equivariant Models
• Image to Icosahedral Projection for SO(3) Object Reasoning from Single-View Images
• Grasp Learning: Models, Methods, and Performance
• BulletArm: An Open-Source Robotic Manipulation Benchmark and Learning Framework
• Edge Grasp Network: A Graph-Based SE(3)-invariant Approach to Grasp Detection
• Leveraging Fully Observable Policies for Learning under Partial Observability
• Visual Foresight With a Local Dynamics Model
• Graph-Structured Policy Learning for Multi-Goal Manipulation Tasks
• Understanding the Mechanism behind Data Augmentation's Success on Image Based RL
• Binding Actions to Objects in World Models
• Equivariant Transporter Network
• SO(2) Equivariant Reinforcement Learning
• Hierarchical Reinforcement Learning under Mixed Observability
• Sample Efficient Grasp Learning Using Equivariant Models
• Binding Actions to Objects in World Models
• Tactile Pose Estimation and Policy Learning for Unknown Object Manipulation
• Policy Learning for Visually Conditioned Tactile Manipulation
• Equivariant Q Learning in Spatial Action Spaces
• GASCN: Graph Attention Shape Completion Network
• Design Guidelines for Human-Robot Interaction with Assistive Robot Manipulation Systems
• Deep Transformer Q-Networks for Partially Observable Reinforcement Learning
• Robotic Pick-and-Place With Uncertain Object Instance Segmentation and Shape Completion
• Action Priors for Large Action Spaces in Robotics
• Efficient and Accurate Candidate Generation for Grasp Pose Detection in SE(3)
• Learning Discrete State Abstractions With Deep Variational Inference
• Learning bayes filter models for tactile localization
• Belief-Grounded Networks for Accelerated Robot Learning under Partial Observability
• Policy learning in SE(3) action spaces
• Learning Manipulation Skills Via Hierarchical Spatial Attention
• Towards Assistive Robotic Pick and Place in Open World Environments
• Learning 6-DoF Grasping and Pick-Place Using Attention Focus
• Online Abstraction with MDP Homomorphisms for Deep Learning
• Learning 6-DoF Grasping and Pick-Place Using Attention Focus
• Pick and Place Without Geometric Object Models
• Learning a visuomotor controller for real world robotic grasping using simulated depth images
• Grasp Pose Detection in Point Clouds
• Open World Assistive Grasping Using Laser Selection
• Viewpoint Selection for Grasp Detection
• High precision grasp pose detection in dense clutter
• Using Geometry to Detect Grasp Poses in 3D Point Clouds.
• Localizing Handle-Like Grasp Affordances in 3D Point Clouds