Robot manipulators achieve 91% accuracy in multi-class human/object detection

Researchers have developed a new method for multi-class human/object detection on robot manipulators, improving upon previous binary classification models. Using a Franka Emika Panda robot, they collected a dataset and trained models like LSTM, GRU, and Transformers. The best-performing model achieved 91.11% accuracy in real-time testing, demonstrating the effectiveness of multi-class detection for more detailed contact analysis in human-robot collaboration. AI

IMPACT Enhances safety and workflow efficiency in physical human-robot collaboration through improved object recognition.

RANK_REASON Academic paper detailing a new methodology and benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Franka Emika Panda
• LSTM
• Maryam Rezayati
• Transformers

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →