GelNeuro: A Sensing-Computing Integrated Neuromorphic Tactile System for Texture Recognition

2026-07-06

AI Summary

This work addresses the limitations of existing neuromorphic tactile systems, which rely on external hosts for event processing and thus struggle to achieve low-power, low-latency edge perception. The study presents the first direct integration of a GelSight Mini optical tactile sensor with the Speck2f neuromorphic system-on-chip to implement an end-to-end event-driven spiking convolutional neural network (SCNN) for texture recognition. By employing a hardware-aware weight pruning strategy, the system achieves enhanced accuracy and improved generalization to unseen indentation depths under 8-bit deployment. Evaluated on a 15-class natural texture recognition task, the on-chip system attains an accuracy of 96.3% within an 80-ms inference window while consuming only 19.6 mW—three orders of magnitude lower power than conventional CPU/GPU-based approaches.

Abstract

Neuromorphic visuo-tactile sensing offers a promising paradigm for low-latency and low-power robotic perception. However, existing systems still rely heavily on a host computer for event readout, preprocessing, or relaying prior to chip inference. This paper presents GelNeuro, a fully integrated sensing-computing visuo-tactile system that directly pairs a GelSight Mini-based optical tactile front end with the Speck2f neuromorphic system-on-chip (SoC). Contact-induced marker motions are captured as dynamic vision sensor (DVS) events and routed through the on-chip network to a spiking convolutional neural network (SCNN) classifier. To mitigate accuracy degradation during 8-bit deployment, a hardware-aware weight clamping strategy is introduced. Evaluated on a 15-class natural texture recognition task, hardware-in-the-loop testing on the physical chip achieves a 96.3% accuracy within an 80 ms inference window. Notably, the system consumes only 19.6 mW of board-level active power-over three orders of magnitude lower than conventional CPU/GPU baselines on the same benchmark. GelNeuro also exhibits robust generalization across unseen contact depths, demonstrating the viability of direct sensor-to-chip tactile recognition on edge neuromorphic hardware.

Problem

• neuromorphic tactile sensing
• texture recognition
• sensing-computing integration
• edge inference
• low-power perception

Innovation

• neuromorphic computing
• sensing-computing integration
• spiking neural network
• tactile sensing
• hardware-aware optimization