VGG16 transfer learning on 1200 building images yields 97.85% training and 89.38% validation accuracy for four-class post-earthquake damage classification.
Performance of low -rise reinforced concrete buildings in the 2016 Ecuador earthquake
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Post-Earthquake Assessment of Buildings Using Deep Learning
VGG16 transfer learning on 1200 building images yields 97.85% training and 89.38% validation accuracy for four-class post-earthquake damage classification.