[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83865-en":3,"doc-seo-83865-105":30,"detail-sidebar-cat-0-en-105":91},{"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},83865,8796095462418,"Noah","https://ap-avatar.wpscdn.com/avatar/80000253c1241d02b47?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778826106357471780",8,"Research & Report","Efficient Perception in Automotive Detection and Tracking Using Neuromorphic Computing","Deep learning for automotive perception faces high energy and compute costs that hinder deployment on edge devices and raise sustainability concerns. Neuromorphic computing with spiking neural networks (SNNs) provides an energy-efficient alternative to von Neumann systems through massively parallel, event-driven processing and on-chip learning. This work evaluates SNNs for real-world multi-object detection and tracking using SpikeYOLO transfer learning, reporting strong KITTI and BDD100K MOT2020 performance competitive with conventional deep learning.","Efficient Perception in Automotive Detection and Tracking Using  \nNeuromorphic Computing  \nManish Kolachalam 1 & Rani Malhotra 1  \n1Applied Research Center for Autonomous Machines, Infosys Center for Emerging Technologies, Bangalore,  \nKarnataka, India, 560100  \nCorresponding author: Manish Kolachalam  \n[Email](Email: manish.k02@infosys.com)[: manish.k02@infosys.com](Email: manish.k02@infosys.com)  \nAbstract  \nDeep learning algorithms are notorious for their high carbon footprint and computational demands that limit their deployment on edge devices and raise concerns about their long-term sustainability. Neuromorphic computing and Spiking Neural Networks (SNNs) offer a promising alternative to traditional Von Neumann architectures, providing energy-efficient performance, massively parallel computation, and on-chip learning capabilities. Autonomous machines represent a critical application domain where these advantages are particularly valuable. We present the first comprehensive evaluation of SNNs for real-world automotive multi-object detection and tracking. Using transfer learning with the SpikeYOLO architecture, we achieve mean Average Precision of 0.937 on the KITTI dataset and 0.771 on BDD100K MOT2020 dataset for object detection and a Higher Order Tracking Accuracy score of 0.701 (KITTI) and 0.445 (BDD100K MOT2020) for object tracking—results competitive with conventional deep learning methods. Our results demonstrate that SNNs can deliver high-performance object detection and tracking in an energyefficient manner, establishing their viability for perception in real-world autonomous systems.  \nKeywords Artificial intelligence, Autonomous systems, Edge AI, Spiking neural networks  \n1 Introduction  \nNeuromorphic computing and Spiking Neural Networks (SNNs) have emerged as a viable alternative to traditional Artificial Neural Networks (ANNs) [1], [2] . These systems are designed to mimic the natural way of information transfer as happens in the brain [3], [4] . While still in its infancy, Neuromorphic systems are the subject of extensive research. Neuromorphic computing allows for energy-efficient and massively parallel data processing due to their asynchronous nature [2], [4] . A variety of different architectures/systems/algorithms have been designed which use biological principles as a guide [2], [5] .  \nThe basic informational unit of neuromorphic computing is the binary spike, modelled on the biological action potential [6] . Energy is only consumed when a neuron emits a spike and not otherwise, as opposed to neurons in ANNs that are always active sequentially and synchronously [5] . Such energy efficient devices would be very relevant to computing on the edge, which requires high efficiency. One such area of interest is autonomous machines like self-driving cars. A self-driving system performs an enormous amount of computation at each time step, based on a whole suite of detectors and requires efficient, fast and parallel infrastructure. Neuromorphic systems provide a compelling alternative to traditional systems with their low energy consumption and highly asynchronous, parallel processing [5] . We are interested  \nin perception in real-world automotive applications using SNNs with the eventual aim of moving these SNNs to specialized neuromorphic hardware like the Intel Loihi 2 [7] or the AKIDA brainchip [8] and deploying them on the edge at scale.  \nA particularly relevant line of research has developed SNNs based on the famous YOLO architecture [9], which provides strong performance on multiple computer vision tasks. These networks, christened SpikeYOLO, have shown strong performance in object detection tasks [10] . Here we validated its applicability for use in autonomous driving systems for the purpose of object detection and tracking, features that are critical for any autonomous vehicle. Multiple factors contribute to making this task harder compared to other benchmarks. Autonomous vehicles function in a dynami","cbCaia5dHM7sc1iw","https://ap.wps.com/l/cbCaia5dHM7sc1iw","pdf",634871,3,1,11,"English","en",105,"# Introduction\n# Related Work\n## Spiking neural networks","[{\"question\":\"Why are neuromorphic computing and SNNs considered alternatives to traditional deep learning for edge deployment?\",\"answer\":\"They are designed to reduce energy use by consuming power primarily when neurons emit spikes, enabling asynchronous, massively parallel processing that suits edge constraints.\"},{\"question\":\"What architecture and training strategy are used in the study?\",\"answer\":\"The study uses transfer learning with the SpikeYOLO architecture to adapt SNN-based models for automotive multi-object detection and tracking.\"},{\"question\":\"How does the proposed approach perform on automotive benchmarks?\",\"answer\":\"On KITTI, it achieves mean Average Precision of 0.937 for detection and Higher Order Tracking Accuracy of 0.701 for tracking; on BDD100K MOT2020, it achieves 0.771 and 0.445 respectively.\"}]",1784191069,28,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"efficient-perception-in-automotive-detection-and-tracking-using-neuromorphic-computing","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":20},"https://docshare.wps.com/document/research-report/",{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/efficient-perception-in-automotive-detection-and-tracking-using-neuromorphic-computing/83865/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why are neuromorphic computing and SNNs considered alternatives to traditional deep learning for edge deployment?","Question",{"text":75,"@type":76},"They are designed to reduce energy use by consuming power primarily when neurons emit spikes, enabling asynchronous, massively parallel processing that suits edge constraints.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What architecture and training strategy are used in the study?",{"text":80,"@type":76},"The study uses transfer learning with the SpikeYOLO architecture to adapt SNN-based models for automotive multi-object detection and tracking.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the proposed approach perform on automotive benchmarks?",{"text":84,"@type":76},"On KITTI, it achieves mean Average Precision of 0.937 for detection and Higher Order Tracking Accuracy of 0.701 for tracking; on BDD100K MOT2020, it achieves 0.771 and 0.445 respectively.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]