activity
20212026
most citedA Computer Vision Enabled damage detection model with improved YOLOv5 based on Transformer Prediction Head

22 citations · 24 across the 4 of their papers we have counts for

collaborators

4 papers