Towards Fast, Memory-based and Data-Efficient Vision-Language Policy
This work addresses efficiency and memory issues in robotic learning with vision-language models, offering an incremental improvement over existing methods.
The paper tackles the challenges of high inference cost, domain shifts, and limited experience handling in vision-language policies for robotics by proposing LiteVLP, a lightweight memory-based model that outperforms state-of-the-art methods on VIMA-Bench with 18.8% improvement in long-horizon tasks and faster inference.
Vision Language Models (VLMs) pretrained on Internet-scale vision-language data have demonstrated the potential to transfer their knowledge to robotic learning. However, the existing paradigm encounters three critical challenges: (1) expensive inference cost resulting from large-scale model parameters, (2) frequent domain shifts caused by mismatched data modalities, and (3) limited capacity to handle past or future experiences. In this work, we propose LiteVLP, a lightweight, memory-based, and general-purpose vision-language policy generation model. LiteVLP is built upon a pre-trained 1B-parameter VLM and fine-tuned on a tiny-scale and conversation-style robotic dataset. Through extensive experiments, we demonstrate that LiteVLP outperforms state-of-the-art vision-language policy on VIMA-Bench, with minimal training time. Furthermore, LiteVLP exhibits superior inference speed while maintaining exceptional high accuracy. In long-horizon manipulation tasks, LiteVLP also shows remarkable memory ability, outperforming the best-performing baseline model by 18.8%. These results highlight LiteVLP as a promising model to integrating the intelligence of VLMs into robotic learning.