Tin Lai

Latest

• A Hybrid Optimization Framework for Grasp Synthesis under Partial Observations
• Diverse Motion Planning with Stein Diffusion Trajectory Inference
• Learning for Kinodynamic Tree Expansion
• Stein Movement Primitives for Adaptive Multi-Modal Trajectory Generation
• Path Signatures for Diversity in Probabilistic Trajectory Optimisation
• LTR*: Rapid Replanning in Consecutive Pick-and-Place Tasks with Lazy Experience Graph
• Learning Efficient and Robust Ordinary Differential Equations via Invertible Neural Networks
• Diffeomorphic Transforms for Generalised Imitation Learning
• Adaptively Exploiting Local Structure with Generalised Multi-Trees Motion Planning
• Parallelised Diffeomorphic Sampling-based Motion Planning
• PlannerFlows: Learning Motion Samplers with Normalising Flows
• Trajectory Generation in New Environments from Past Experiences
• Anticipatory Navigation in Crowds by Probabilistic Prediction of Pedestrian Future Movements