In offline reinforcement learning (RL), the performance of the learned policy highly depends on the quality of offline datasets. However, in many cases, the offline dataset contains very limited optimal trajectories, which poses a challenge for offline RL algorithms as agents must acquire the ability to transit to high-reward regions. To address this issue, we introduce Diffusion-based Trajectory Stitching (DiffStitch), a novel diffusion-based data augmentation pipeline that systematically generates stitching transitions between trajectories. DiffStitch effectively connects low-reward trajectories with high-reward trajectories, forming globally optimal trajectories to address the challenges faced by offline RL algorithms. Empirical experiments conducted on D4RL datasets demonstrate the effectiveness of DiffStitch across RL methodologies. Notably, DiffStitch demonstrates substantial enhancements in the performance of one-step methods (IQL), imitation learning methods (TD3+BC), and trajectory optimization methods (DT).

离线强化学习中，学到的策略性能高度依赖于离线数据集的质量。然而，在许多情况下，离线数据集只包含非常有限的最优轨迹，这对离线强化学习算法提出了挑战，因为智能体必须能够转移到高奖励区域。为了解决这个问题，我们引入了一种基于扩散的轨迹拼接（DiffStitch）方法，这是一种新颖的基于扩散的数据增强技术，能够在轨迹之间生成连接转移。DiffStitch有效地连接了低奖励轨迹与高奖励轨迹，形成全局最优轨迹，以解决离线强化学习算法面临的挑战。在D4RL数据集上进行的实证实验表明DiffStitch在RL方法中的有效性。值得注意的是，DiffStitch在一步方法（IQL）、模仿学习方法（TD3+BC）和轨迹优化方法（DT）的性能上有了显著提升。

DiffStitch：基于扩散的轨迹拼接技术提升离线强化学习