Diffusion-based Planning with Learned Viability Filters

TL;DR

This paper introduces learned viability filters that predict the success of diffusion-based motion plans, enabling efficient, constraint-aware planning for complex 3D human locomotion tasks.

Contribution

It presents a novel approach to enforce implicit constraints in diffusion models using learned viability filters, improving planning efficiency and success in challenging environments.

Findings

Viability filters effectively predict plan success.

The approach enables online planning and control.

It is significantly faster than guidance-based methods.

Abstract

Diffusion models can be used as a motion planner by sampling from a distribution of possible futures. However, the samples may not satisfy hard constraints that exist only implicitly in the training data, e.g., avoiding falls or not colliding with a wall. We propose learned viability filters that efficiently predict the future success of any given plan, i.e., diffusion sample, and thereby enforce an implicit future-success constraint. Multiple viability filters can also be composed together. We demonstrate the approach on detailed footstep planning for challenging 3D human locomotion tasks, showing the effectiveness of viability filters in performing online planning and control for box-climbing, step-over walls, and obstacle avoidance. We further show that using viability filters is significantly faster than guidance-based diffusion prediction.