Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders.
Paul BergmannSindy LöweMichael FauserDavid SattleggerCarsten StegerPublished in: VISIGRAPP (5: VISAPP) (2019)
Keyphrases
- structural similarity
- structural information
- fully unsupervised
- image quality assessment
- denoising
- correlation coefficient
- unsupervised learning
- image segmentation
- multiscale
- quality metrics
- supervised learning
- co occurrence
- structural similarity index
- knowledge base
- precision and recall
- high frequency
- image regions
- low level
- data analysis