[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-31264":3,"doc-seo-31264":26},{"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,"file_id":15,"file_url":16,"file_type":17,"file_size":18,"view_count":19,"is_deleted":4,"is_public":19,"is_downloadable":19,"audit_status":19,"page_count":20,"language":21,"table_of_contents":22,"faqs":23,"seo_title":13,"seo_description":14,"update_tm":24,"read_time":25},31264,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",6,"Technology","PHL YOLO 实时轻量化纱线检测方法","A real-time, high-precision yarn inspection approach is presented to meet practical production needs. The method, PHL-YOLO, is built on an optimized YOLOv8 framework by introducing the PEC2F module to boost feature extraction while lowering computation and parameters, and the HWD module to downsample images without losing key information. To reduce detection-head cost, an LSCDH lightweight shared convolutional head is designed and a structured pruning strategy (GSPMD with DepGraph) further streamlines the network. Experiments show 1.0% higher precision, 2.5% better recall, and 2.3% improved mAP, plus a 26 FPS gain, with computational load and parameters at 29.6% and 10.2% of original YOLOv8.","cbCaiqIjWes33SiI","https://ap.wps.com/l/cbCaiqIjWes33SiI","pdf",2020104,1,14,"English","# 1 Introduction\n# 2 Model development\n## 2.1 Benchmark model: YOLOv8","[{\"question\":\"What problem does PHL-YOLO target in yarn production?\",\"answer\":\"It targets the need for real-time and high-precision yarn anomaly/quality detection during actual yarn manufacturing, where manual inspection is labor-intensive and expertise-dependent.\"},{\"question\":\"How does PHL-YOLO improve the original YOLOv8 architecture?\",\"answer\":\"It adds the PEC2F module for stronger feature extraction with lower computation, uses HWD for efficient downsampling, replaces the YOLOv8 detection head with LSCDH to simplify computation, and applies GSPMD structured pruning to further reduce model complexity.\"},{\"question\":\"What experimental gains are reported compared with the original YOLOv8?\",\"answer\":\"The model achieves a 1.0% increase in precision, a 2.5% improvement in recall, and a 2.3% rise in mAP, alongside a 26 FPS performance boost; computational load and parameters are reduced to 29.6% and 10.2%, respectively.\"}]",1779224602,35,{"code":4,"msg":27,"data":28},"ok",{"site_id":29,"language":30,"slug":31,"title":13,"keywords":32,"description":14,"schema_data":33,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":24},105,"en","phl-yolo-real-time-lightweight-yarn-inspection-method","",{"@graph":34,"@context":84},[35,52,67],{"@type":36,"itemListElement":37},"BreadcrumbList",[38,42,46,49],{"item":39,"name":40,"@type":41,"position":19},"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/technology/",3,{"item":50,"name":13,"@type":41,"position":51},"https://docshare.wps.com/document/phl-yolo-real-time-lightweight-yarn-inspection-method/31264/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"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","2026-05-20","2026-05-19",true,{"@type":64,"interactionType":65,"userInteractionCount":19},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does PHL-YOLO target in yarn production?","Question",{"text":74,"@type":75},"It targets the need for real-time and high-precision yarn anomaly/quality detection during actual yarn manufacturing, where manual inspection is labor-intensive and expertise-dependent.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does PHL-YOLO improve the original YOLOv8 architecture?",{"text":79,"@type":75},"It adds the PEC2F module for stronger feature extraction with lower computation, uses HWD for efficient downsampling, replaces the YOLOv8 detection head with LSCDH to simplify computation, and applies GSPMD structured pruning to further reduce model complexity.",{"name":81,"@type":72,"acceptedAnswer":82},"What experimental gains are reported compared with the original YOLOv8?",{"text":83,"@type":75},"The model achieves a 1.0% increase in precision, a 2.5% improvement in recall, and a 2.3% rise in mAP, alongside a 26 FPS performance boost; computational load and parameters are reduced to 29.6% and 10.2%, respectively.","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":29}]