Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
arXiv:2110.03659
Abstract
An agent's functionality is largely determined by its design, i.e., skeletal structure and joint attributes (e.g., length, size, strength). However, finding the optimal agent design for a given function is extremely challenging since the problem is inherently combinatorial and the design space is prohibitively large. Additionally, it can be costly to evaluate each candidate design which requires solving for its optimal controller. To tackle these problems, our key idea is to incorporate the design procedure of an agent into its decision-making process. Specifically, we learn a conditional policy that, in an episode, first applies a sequence of transform actions to modify an agent's skeletal structure and joint attributes, and then applies control actions under the new design. To handle a variable number of joints across designs, we use a graph-based policy where each graph node represents a joint and uses message passing with its neighbors to output joint-specific actions. Using policy gradient methods, our approach enables joint optimization of agent design and control as well as experience sharing across different designs, which improves sample efficiency substantially. Experiments show that our approach, Transform2Act, outperforms prior methods significantly in terms of convergence speed and final performance. Notably, Transform2Act can automatically discover plausible designs similar to giraffes, squids, and spiders. Code and videos are available at https://sites.google.com/view/transform2act.
ICLR 2022 (Oral). Project page: https://sites.google.com/view/transform2act. Code: https://github.com/Khrylx/Transform2Act
References in corpus (9)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Semi-Supervised Classification with Graph Convolutional Networks
- Fast Graph Representation Learning with PyTorch Geometric
- Neural Graph Evolution: Towards Efficient Automatic Robot Design
- One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Control
- Contrasting Exploration in Parameter and Action Space: A Zeroth-Order Optimization Perspective
- SimGAN: Hybrid Simulator Identification for Domain Adaptation via Adversarial Reinforcement Learning
- Hardware as Policy: Mechanical and Computational Co-Optimization using Deep Reinforcement Learning
- Task-Agnostic Morphology Evolution