Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
March 23, 2026Biomimetic Intelligence and RoboticsOpen Access

Enhancing stability and reliability in LLM-driven robotic manipulation through human skill demonstration and visual tracking

View Full Paper
Ask AI
Bookmark
Share

Authors

GYGuo YiHCHao ChuSWShiguang Wen

Discussion

Loading...

Member takes

Overview

Novel framework enhances robotic skill transfer and decision-making through human input and environmental feedback.

Key Points

  • The aim is to improve robotic manipulation stability and reliability using LLMs and human skill demonstrations.
  • Developed an LLM-driven framework for robotic operations.
  • Implemented Stable chain-of-thought and prompt engineering for multi-step reasoning.
  • Utilized Beta Process Autoregressive Hidden Markov Model for skill segmentation.
  • Integrated Dynamic Movement Primitives for anthropomorphic operation learning.
  • Incorporated a visual foundation model for environmental perception.
  • Framework significantly outperformed baseline methods in decision accuracy.
  • Achieved improved reasoning stability and adaptability.
  • Enabled complex trajectory tasks and effective skill generalization.
  • Confirmed through both simulations and real-world experiments.

Cite This Study

Yi et al. (2026) studied this question.

synapsesocial.com/papers/69c0df0bfddb9876e79c1503https://doi.org/10.1016/j.birob.2026.100305
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Agentic Skill Discovery2024
  2. 2Enhancing the LLM-Based Robot Manipulation Through Human-Robot Collaboration2024 · 92 citations
  3. 3Lifelong Robot Library Learning: Bootstrapping Composable and Generalizable Skills for Embodied Control with Language Models2024 · 1 citations
  4. 4Towards Natural Language-Driven Assembly Using Foundation Models2024
  5. 5Empowering Large Language Models on Robotic Manipulation with Affordance Prompting2024 · 4 citations