[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84802-en":3,"doc-seo-84802-105":28,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":13,"seo_description":14,"update_tm":26,"read_time":27},84802,5909877438554,"Maeve","https://ap-avatar.wpscdn.com/avatar/5600025385ad2bf12a7?_k=1778553567797529272",8,"Research & Report","GelNeuro A Sensing Computing Integrated Neuromorphic Tactile System for Texture Recognition","Neuromorphic visuo-tactile sensing enables low-latency, low-power robotic perception, but many existing designs depend on a host computer for event readout, preprocessing, or relaying to the chip. The work introduces GelNeuro, integrating a GelSight Mini optical tactile front end with the Speck2f neuromorphic SoC. DVS event streams from contact-induced marker motions are routed on-chip to a spiking convolutional neural network classifier. Hardware-aware weight clamping supports stable 8-bit deployment, achieving 96.3% accuracy on a 15-class texture task in an 80 ms window with 19.6 mW power.","GelNeuro: A Sensing-Computing Integrated Neuromorphic Tactile  \nSystem for Texture Recognition  \nLuoyang Bian, Xinpan Meng, Zhenghua Ma, Houcheng Li and Long Cheng, Fellow, IEEE  \narXiv :2607 .0524 1v 1 [ cs .RO] 6 Jul 2026  \nAbstract—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 sensingcomputing 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 boardlevel 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.  \nIndex Terms—Neuromorphic tactile perception, in-sensor computing, spiking convolutional neural network (SCNN), visuotactile sensor, texture recognition.  \nI. INTRODUCTION  \nTACTILE sensing provides high-fidelity local contact in  \nformation, including fine surface geometry and incipient slip, that is inherently difficult to infer from vision alone yet remains essential for dexterous robotic manipulation in cluttered or occluded environments [1] . Texture recognition serves as a rigorous benchmark for tactile systems because it simultaneously demands high spatial resolution to capture intricate surface profiles and exceptional temporal resolution to resolve transient contact dynamics [2] . When incorporating tactile feedback into closed-loop control, both latency and operating power directly dictate the available reaction window and the viability of edge deployment on resource-constrained platforms [3] . While camera-based visuo-tactile sensors offer rich spatial resolution by imaging the deformation of an  \nThis work was supported in part by the Brain Science and Brain-like Intelligence Technology - National Science and Technology Major Project under Grant 2025ZD0215600, in part by the National Natural Science Foundation of China under Grant U25A20475, in part by the Beijing Municipal Natural Science Foundation under Grants F2024201068, L243014 and L232140, in part by Young Scientists Fund of The State Key Laboratory of Multimodal Artificial Intelligence Systems under Grant ES2P100114, and in part by the Fundamental Research Funds for the Central Universities. (Corresponding author: Houcheng Li and Long Cheng.)  \nThe authors are all with the State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China, and also with the School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China. All correspondences should be addressed to the corresponding author Prof. Long Cheng([long.cheng@ia.ac.cn](long.cheng@ia.ac.cn)).  \n\n| (a) Visuo-tactile Sensor\u003Cbr>\u003Cbr>Raw visuotactile events |  | (b) Speck2fNeuromorphic System-on-Chip\u003Cbr>\u003Cbr>On-Chip SNN inference |\n| --- | --- | --- |\n\nFig. 1. GelNeuro tactile perception pipeline. (a) A visuo-tactile sensor is directly interfaced with the Speck2f neuromorphic SoC, which captures raw event frames accumulated from marker motions on the gel surface. (b) The Speck2f neuromorphic SoC handles edge computing, executing on-c","cbCaidXrxXpXXPDT","https://ap.wps.com/l/cbCaidXrxXpXXPDT","pdf",1691027,1,"English","en",105,"# Abstract\n# Index Terms\n# Introduction\n## Tactile sensing and texture recognition\n## Limitations of frame-based and host-dependent pipelines\n## Event-based sensing and neuromorphic processing","[{\"question\":\"What problem does GelNeuro address in existing neuromorphic visuo-tactile systems?\",\"answer\":\"Existing systems often rely on a host computer for event readout, preprocessing, or relaying event streams to the neuromorphic chip, which keeps latency and power benefits limited.\"},{\"question\":\"How does GelNeuro connect the tactile sensor to neuromorphic inference?\",\"answer\":\"GelSight Mini captures optical tactile inputs, which are converted into DVS events from marker motions and routed through the Speck2f on-chip network to a spiking convolutional neural network classifier.\"},{\"question\":\"How is accuracy maintained when deploying the model on 8-bit hardware?\",\"answer\":\"GelNeuro introduces a hardware-aware weight clamping strategy to mitigate accuracy degradation during 8-bit deployment.\"}]",1784198339,20,{"code":4,"msg":29,"data":30},"ok",{"site_id":23,"language":22,"slug":31,"title":13,"keywords":32,"description":14,"schema_data":33,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":26},"gelneuro-a-sensing-computing-integrated-neuromorphic-tactile-system-for-texture-recognition","",{"@graph":34,"@context":84},[35,52,67],{"@type":36,"itemListElement":37},"BreadcrumbList",[38,42,46,49],{"item":39,"name":40,"@type":41,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":43,"name":44,"@type":41,"position":45},"https://docshare.wps.com/document/","Document",2,{"item":47,"name":12,"@type":41,"position":48},"https://docshare.wps.com/document/research-report/",3,{"item":50,"name":13,"@type":41,"position":51},"https://docshare.wps.com/document/gelneuro-a-sensing-computing-integrated-neuromorphic-tactile-system-for-texture-recognition/84802/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"inLanguage":22,"description":14,"dateModified":60,"datePublished":61,"encodingFormat":59,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":55},"Person",{"url":39,"name":57,"@type":58},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":20},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does GelNeuro address in existing neuromorphic visuo-tactile systems?","Question",{"text":74,"@type":75},"Existing systems often rely on a host computer for event readout, preprocessing, or relaying event streams to the neuromorphic chip, which keeps latency and power benefits limited.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does GelNeuro connect the tactile sensor to neuromorphic inference?",{"text":79,"@type":75},"GelSight Mini captures optical tactile inputs, which are converted into DVS events from marker motions and routed through the Speck2f on-chip network to a spiking convolutional neural network classifier.",{"name":81,"@type":72,"acceptedAnswer":82},"How is accuracy maintained when deploying the model on 8-bit hardware?",{"text":83,"@type":75},"GelNeuro introduces a hardware-aware weight clamping strategy to mitigate accuracy degradation during 8-bit deployment.","https://schema.org",{"og:url":50,"og:type":86,"og:title":13,"og:site_name":57,"og:description":14},"article",{"robots":88,"canonical":50},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":44,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":45,"doc_module":4,"doc_module_name":44,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":51,"doc_module":4,"doc_module_name":44,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":44,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":44,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":44,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":44,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":44,"category_name":124,"show_sort_weight":27,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":27,"doc_module":4,"doc_module_name":44,"category_name":127,"show_sort_weight":27,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":44,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":44,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]