Lightweight CNN deciphers affective touch in soft companions

Researchers have developed a lightweight 1D Convolutional Neural Network (CNN) for classifying affective touch in soft robotic companions. This study introduces an open-source MATLAB framework and a dataset of 1326 labeled gesture sequences from 25 participants. The compact CNN model, with 13.2k parameters, achieved 75% test accuracy and is capable of real-time operation on microcontrollers, enabling privacy-preserving touch interpretation within therapeutic devices. AI

IMPACT Enables more nuanced and privacy-preserving human-robot interaction in therapeutic soft companions.

RANK_REASON The cluster describes an academic paper detailing a novel model and dataset for a specific application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• 1D CNN
• Aleksandrs Vališevskis
• CatalyzeX
• CNN
• DagsHub
• Hugging Face
• MATLAB

• paper
• product
• infra

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