Stochastic Actor-Executor-Critic for Image-to-Image Translation

TL;DR

This paper introduces the Stochastic Actor-Executor-Critic (SAEC), a novel reinforcement learning framework that effectively handles high-dimensional image-to-image translation tasks by combining stochastic policies with an executor for realistic image generation.

Contribution

The paper proposes the SAEC model, integrating stochastic policies and an executor within an off-policy actor-critic framework for improved image translation.

Findings

SAEC outperforms existing methods on multiple image translation benchmarks.

The approach demonstrates robustness in high-dimensional continuous spaces.

Experimental results validate the effectiveness of the stochastic actor and executor design.

Abstract

Training a model-free deep reinforcement learning model to solve image-to-image translation is difficult since it involves high-dimensional continuous state and action spaces. In this paper, we draw inspiration from the recent success of the maximum entropy reinforcement learning framework designed for challenging continuous control problems to develop stochastic policies over high dimensional continuous spaces including image representation, generation, and control simultaneously. Central to this method is the Stochastic Actor-Executor-Critic (SAEC) which is an off-policy actor-critic model with an additional executor to generate realistic images. Specifically, the actor focuses on the high-level representation and control policy by a stochastic latent action, as well as explicitly directs the executor to generate low-level actions to manipulate the state. Experiments on several image-to-image translation tasks have demonstrated the effectiveness and robustness of the proposed SAEC when facing high-dimensional continuous space problems.