Generative robot policies such as Flow Matching offer flexible, multi-modal policy learning but are sample-inefficient. Although object-centric policies improve sample efficiency, it does not resolve this limitation. In this work, we propose Multi-Stream Generative Policy (MSG), an inference-time composition framework that trains multiple object-centric policies and combines them at inference to improve generalization and sample efficiency. MSG is model-agnostic and inference-only, hence widely applicable to various generative policies and training paradigms. We perform extensive experiments both in simulation and on a real robot, demonstrating that our approach learns high-quality generative policies from as few as five demonstrations, resulting in a 95% reduction in demonstrations, and improves policy performance by 89 percent compared to single-stream approaches. Furthermore, we present comprehensive ablation studies on various composition strategies and provide practical recommendations for deployment. Finally, MSG enables zero-shot object instance transfer. We make our code publicly available at https://msg.cs.uni-freiburg.de.

翻译：生成式机器人策略（如流匹配）提供了灵活的多模态策略学习能力，但存在样本效率低下的问题。尽管以物体为中心的策略能提升样本效率，但并未从根本上解决这一局限性。本文提出多流生成策略（MSG），这是一种推理时组合框架，通过训练多个以物体为中心的策略并在推理阶段进行组合，以提升泛化能力与样本效率。MSG具有模型无关性且仅需推理时操作，因此可广泛适用于各类生成式策略与训练范式。我们在仿真环境与真实机器人上开展了大量实验，结果表明：该方法仅需5条示范数据即可学习到高质量的生成式策略，示范数据需求量降低95%，且策略性能相比单流方法提升89%。此外，我们针对不同组合策略进行了全面的消融研究，并为实际部署提供了实用建议。最终，MSG实现了零样本物体实例迁移。相关代码已开源至https://msg.cs.uni-freiburg.de。