A context-aware generative model synthesizes crack growth patterns with directional propagation and learned morphology to augment training data and improve crack segmentation performance.
A new gan- based approach to data augmentation and image segmentation for crack detection in thermal imaging tests
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CrackForward: Context-Aware Severity Stage Crack Synthesis for Data Augmentation
A context-aware generative model synthesizes crack growth patterns with directional propagation and learned morphology to augment training data and improve crack segmentation performance.