Comparison between the improved model and the original YOLOv8 classification model.

<p>Comparison between the improved model and the original YOLOv8 classification model.</p>

Furkejuvvon:

Bibliográfalaš dieđut
Váldodahkki:	Qingtao Zeng (6270308) (author)
Eará dahkkit:	Lixin Zhang (55947) (author), Jiefeng Zhao (9090305) (author), Anping Xu (5729921) (author), Yali Qi (8375922) (author), Liqin Yu (1461880) (author), Wenjing Li (198063) (author), Haochang Xia (22582017) (author)
Almmustuhtton:	2025
Fáttát:	Space Science Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified traditional upsample method real datasets validate managing computational costs dynamic sampling method capture intricate features achieve automatic identification 0 %, respectively yield superior performance original yolov8 framework learning key features top_5 accuracy </ top_1 accuracy </ scale attention mechanism point convolution within integrates parallel convolution div >< p book cover classification scale attention performance enhancement dualconv framework deep learning book covers study introduces significantly enhance pwconv operators precise classification novel dual management efficiency inherent limitations image classification head section experimental results employed instead efficient multi computer vision books published backbone network approach overcomes annual increase
Fáddágilkorat:	Lasit fáddágilkoriid Eai fáddágilkorat, Lasit vuosttaš fáddágilkora!

Geahča maid