VLMimic: Vision Language Models are Visual Imitation Learner for Fine-grained Actions

TL;DR

VLMimic leverages vision language models to directly learn fine-grained robotic actions from limited human videos, enabling efficient adaptation and significant performance improvements in manipulation tasks.

Contribution

The paper introduces VLMimic, a novel approach that uses VLMs for fine-grained action learning directly from limited videos, bypassing high-level plan abstraction.

Findings

Achieves over 27% improvement in RLBench tasks with only 5 videos.

Surpasses baselines by over 37% in long-horizon tasks.

Demonstrates effective adaptation to unseen environments.

Abstract

Visual imitation learning (VIL) provides an efficient and intuitive strategy for robotic systems to acquire novel skills. Recent advancements in Vision Language Models (VLMs) have demonstrated remarkable performance in vision and language reasoning capabilities for VIL tasks. Despite the progress, current VIL methods naively employ VLMs to learn high-level plans from human videos, relying on pre-defined motion primitives for executing physical interactions, which remains a major bottleneck. In this work, we present VLMimic, a novel paradigm that harnesses VLMs to directly learn even fine-grained action levels, only given a limited number of human videos. Specifically, VLMimic first grounds object-centric movements from human videos, and learns skills using hierarchical constraint representations, facilitating the derivation of skills with fine-grained action levels from limited human videos. These skills are refined and updated through an iterative comparison strategy, enabling efficient adaptation to unseen environments. Our extensive experiments exhibit that our VLMimic, using only 5 human videos, yields significant improvements of over 27% and 21% in RLBench and real-world manipulation tasks, and surpasses baselines by over 37% in long-horizon tasks.