Gated Path Planning Networks

TL;DR

This paper introduces the Gated Path Planning Network, an improved differentiable navigation module that addresses optimization issues in Value Iteration Networks by incorporating gating mechanisms, leading to better performance and generalization.

Contribution

It reframes VINs as gated recurrent convolutional networks, proposing a new architecture that improves training stability and generalization in path planning tasks.

Findings

Outperforms VIN in learning speed and stability

Shows robustness across maze types and sizes

Succeeds in a 3D environment with RGB inputs

Abstract

Value Iteration Networks (VINs) are effective differentiable path planning modules that can be used by agents to perform navigation while still maintaining end-to-end differentiability of the entire architecture. Despite their effectiveness, they suffer from several disadvantages including training instability, random seed sensitivity, and other optimization problems. In this work, we reframe VINs as recurrent-convolutional networks which demonstrates that VINs couple recurrent convolutions with an unconventional max-pooling activation. From this perspective, we argue that standard gated recurrent update equations could potentially alleviate the optimization issues plaguing VIN. The resulting architecture, which we call the Gated Path Planning Network, is shown to empirically outperform VIN on a variety of metrics such as learning speed, hyperparameter sensitivity, iteration count, and even generalization. Furthermore, we show that this performance gap is consistent across different maze transition types, maze sizes and even show success on a challenging 3D environment, where the planner is only provided with first-person RGB images.