A PCA-guided autoencoder with a distillation loss is claimed to give a more structured latent space and better defect characterization than existing dimensionality-reduction methods in active infrared thermography.
The results in Table 7 indicate that: (1) The segment partition has a minor effect on evaluation, as the model is trained with various partition strategies
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
PCA-Guided Autoencoding for Structured Dimensionality Reduction in Active Infrared Thermography
A PCA-guided autoencoder with a distillation loss is claimed to give a more structured latent space and better defect characterization than existing dimensionality-reduction methods in active infrared thermography.