PulseTrendingJournal ClubResearchersJournalsExplore
Instagram
HomeTrendingJournal ClubExplore
Synapse
⌘+K
Synapse
March 21, 2026The International Journal of Robotics ResearchOpen Access

Enhancing interpretability and interactivity in robot manipulation: A neurosymbolic approach

View Full Paper
Ask AI
Bookmark
Share

Authors

GTGeorgios TziafasHKH. Kasaei

Discussion

Loading...

Member takes

Overview

Neurosymbolic architecture improves robot manipulation accuracy in interactive tasks, suggesting enhanced user engagement.

Key Points

  • This research aims to enhance robot manipulation through a neurosymbolic system that integrates natural language processing and visual reasoning.
  • Developed a neurosymbolic architecture connecting language and visual reasoning.
  • Created a library of primitive skills for task-agnostic manipulation.
  • Generated a synthetic dataset of 3D scenes for model training.
  • Evaluated the system's performance in both synthetic and real-world environments.
  • Achieved an average success rate of 80.2% in interactive object-picking tasks.
  • Demonstrated robustness to user vocabulary variations.
  • Showcased benefits in accuracy and sample efficiency through few-shot learning.

Cite This Study

Tziafas et al. (2026) studied this question.

synapsesocial.com/papers/69be38ca6e48c4981c679667https://doi.org/10.1177/02783649251412867
View Full Paper
Ask AI
Bookmark
Share