On an augmented 400-image concrete damage dataset, YOLO-v7 instance segmentation reached 96.1% mAP@0.5 at 40 FPS, outperforming Mask R-CNN at 92.1% mAP@0.5 and 18 FPS.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Data-driven Detection and Evaluation of Damages in Concrete Structures: Using Deep Learning and Computer Vision
On an augmented 400-image concrete damage dataset, YOLO-v7 instance segmentation reached 96.1% mAP@0.5 at 40 FPS, outperforming Mask R-CNN at 92.1% mAP@0.5 and 18 FPS.