朝向協同、通用和高效的雙系統機器人操作模式前進

Towards Synergistic, Generalized, and Efficient Dual-System for Robotic Manipulation

October 10, 2024

作者: Qingwen Bu, Hongyang Li, Li Chen, Jisong Cai, Jia Zeng, Heming Cui, Maoqing Yao, Yu Qiao

cs.AI

摘要

隨著對多功能機器人系統在各種多變環境中運作的需求不斷增加，強調了通用政策的重要性，這種政策利用大量跨體驗數據語料庫來促進廣泛適應性和高層次推理。然而，通用政策在推理效率和訓練成本昂貴方面存在困難。相反，專家政策是針對特定領域數據精心製作的，在任務級別精度和效率方面表現出色。然而，它缺乏對各種應用的泛化能力。受到這些觀察的啟發，我們介紹了RoboDual，這是一種協同雙系統，補充了通用政策和專家政策的優點。基於擴散變壓器的專家被設計用於多步驟動作展開，精心條件化於高層次任務理解和視覺-語言-動作（VLA）通用政策的離散動作輸出。與OpenVLA相比，RoboDual 在現實世界環境中實現了26.7% 的改進，在CALVIN上則提高了12%，僅通過引入具有僅2000萬可訓練參數的專家政策。它僅使用5% 的示範數據即可保持強大性能，在現實世界部署中實現了3.8倍更高的控制頻率。代碼將公開提供。我們的項目頁面位於：https://opendrivelab.com/RoboDual/

English

The increasing demand for versatile robotic systems to operate in diverse and dynamic environments has emphasized the importance of a generalist policy, which leverages a large cross-embodiment data corpus to facilitate broad adaptability and high-level reasoning. However, the generalist would struggle with inefficient inference and cost-expensive training. The specialist policy, instead, is curated for specific domain data and excels at task-level precision with efficiency. Yet, it lacks the generalization capacity for a wide range of applications. Inspired by these observations, we introduce RoboDual, a synergistic dual-system that supplements the merits of both generalist and specialist policy. A diffusion transformer-based specialist is devised for multi-step action rollouts, exquisitely conditioned on the high-level task understanding and discretized action output of a vision-language-action (VLA) based generalist. Compared to OpenVLA, RoboDual achieves 26.7% improvement in real-world setting and 12% gain on CALVIN by introducing a specialist policy with merely 20M trainable parameters. It maintains strong performance with 5% of demonstration data only, and enables a 3.8 times higher control frequency in real-world deployment. Code would be made publicly available. Our project page is hosted at: https://opendrivelab.com/RoboDual/