pone.0289179.g006.tif (5.47 MB)
The prediction results of deep learning algorithm for defect data sets.
figure
posted on 2023-12-07, 18:39 authored by Yuntao Xu, Peigang Jiao, Jiaqi LIU(a) The traditional YOLOv5 prediction box for defect types. (b) CFM-YOLOv5 algorithm for defect type prediction box.
History
Usage metrics
Categories
Keywords
xlink "> aimingstandard transformer encodermetal surface defectsfeature extraction abilityexperimental results showgreatly improved mapfeature enhancement partimproved yolov5 algorithmimproved algorithmprediction partreference significancenegative impactlow efficiencyhead selfevc moduledata setcfpnet moudleattention moduleanalogy experimentsaccurately identifyablation experiments
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC