Top 5 Best-performing and top 5 least-performing classes for the simple ViT.

<p>Top 5 Best-performing and top 5 least-performing classes for the simple ViT.</p>

Saved in:

Bibliographic Details
Main Author:	Isra Naz (22560205) (author)
Other Authors:	Jamal Hussain Shah (16896345) (author), Ali Tahir (4342723) (author), Mahatma Reddy Marri (22560208) (author), Rabia Saleem (13802413) (author), Mutaz Elradi S. Saeed (22560211) (author)
Published:	2025
Subjects:	Space Science Biological Sciences not elsewhere classified Information Systems not elsewhere classified varying feature extraction targets specific types subpar feature extraction models still rely main component consists input sign type high classification performance fixed projection matrices adaptive gating technique macro f1 score fully obscured signs assisted driving technologies traffic sign detection conditional visual transformer 07 %, respectively traffic sign recognition static attention mechanisms proposed cvit achieved traffic signs 07 %, object detection macro recall driving vehicles conditional multi vision transformer head transformer 87 %, 3 %, xlink "> weather conditions v projections v ). results demonstrate recent advances overall accuracy optimally adjusts misclassification problem minimizing misclassifications micro precision mechanism limits handle real extensively studied based mechanisms attention adjustments
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items