Learning Neural Parsers with Deterministic Differentiable Imitation Learning

TL;DR

This paper introduces a neural parsing method for spatial tasks, inspired by decision trees, trained with a novel deterministic policy gradient technique called DRAG, enabling better generalization without ground-truth annotations.

Contribution

The paper proposes a deterministic policy gradient-based imitation learning approach for neural parsers, improving performance over existing methods in spatial task decomposition.

Findings

Outperforms state-of-the-art imitation learning methods

Effective in parsing natural images without ground truth

Introduces the DRAG algorithm for deterministic policy updates

Abstract

We explore the problem of learning to decompose spatial tasks into segments, as exemplified by the problem of a painting robot covering a large object. Inspired by the ability of classical decision tree algorithms to construct structured partitions of their input spaces, we formulate the problem of decomposing objects into segments as a parsing approach. We make the insight that the derivation of a parse-tree that decomposes the object into segments closely resembles a decision tree constructed by ID3, which can be done when the ground-truth available. We learn to imitate an expert parsing oracle, such that our neural parser can generalize to parse natural images without ground truth. We introduce a novel deterministic policy gradient update, DRAG (i.e., DeteRministically AGgrevate) in the form of a deterministic actor-critic variant of AggreVaTeD, to train our neural parser. From another perspective, our approach is a variant of the Deterministic Policy Gradient suitable for the imitation learning setting. The deterministic policy representation offered by training our neural parser with DRAG allows it to outperform state of the art imitation and reinforcement learning approaches.