Dynamics Learning with Cascaded Variational Inference for Multi-Step Manipulation

Abstract

The fundamental challenge of planning for multi-step manipulation is to find effective and plausible action sequences that lead to the task goal. We present Cascaded Variational Inference (CAVIN) Planner, a model-based method that hierarchically generates plans by sampling from latent spaces. To facilitate planning over long time horizons, our method learns latent representations that decouple the prediction of high-level effects from the generation of low-level motions through cascaded variational inference. This enables us to model dynamics at two different levels of temporal resolutions for hierarchical planning. We evaluate our approach in three multi-step robotic manipulation tasks in cluttered tabletop environments given high-dimensional observations. Empirical results demonstrate that the proposed method outperforms state-of-the-art model-based methods by strategically interacting with multiple objects.

CoRL 2019

References in corpus (1)

• Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddings

Cited by in corpus (8)

• PlaTe: Visually-Grounded Planning with Transformers in Procedural Tasks
• Long-Horizon Visual Planning with Goal-Conditioned Hierarchical Predictors
• Deep Imitation Learning for Bimanual Robotic Manipulation
• 3D-OES: Viewpoint-Invariant Object-Factorized Environment Simulators
• Dreaming: Model-based Reinforcement Learning by Latent Imagination without Reconstruction
• A Long Horizon Planning Framework for Manipulating Rigid Pointcloud Objects
• Adaptive Procedural Task Generation for Hard-Exploration Problems
• Subgoal Search For Complex Reasoning Tasks