Mamba as a motion encoder for robotic imitation learning

TL;DR

This paper proposes using Mamba, a state-of-the-art architecture with potential applications in LLMs, for robotic imitation learning, highlighting its ability to function as an encoder that effectively captures contextual information by reducing the dimensionality of the state space.

Abstract

Recent advancements in imitation learning, particularly with the integration of LLM techniques, are set to significantly improve robots' dexterity and adaptability. This paper proposes using Mamba, a state-of-the-art architecture with potential applications in LLMs, for robotic imitation learning, highlighting its ability to function as an encoder that effectively captures contextual information. By reducing the dimensionality of the state space, Mamba operates similarly to an autoencoder. It effectively compresses the sequential information into state variables while preserving the essential temporal dynamics necessary for accurate motion prediction. Experimental results in tasks such as cup placing and case loading demonstrate that despite exhibiting higher estimation errors, Mamba achieves superior success rates compared to Transformers in practical task execution. This performance is attributed to Mamba's structure, which encompasses the state space model. Additionally, the study investigates Mamba's capacity to serve as a real-time motion generator with a limited amount of training data.

Figures (8)

• Figure 1: Architecture of the proposed model.
• Figure 2: Mamba block.
• Figure 3: Training using bilateral control system.
• Figure 4: Time responses of variables in state space.
• Figure 5: RMSE of fixed A matrices.