Matches in SemOpenAlex for { <https://semopenalex.org/work/W3197758081> ?p ?o ?g. }
Showing items 1 to 60 of
60
with 100 items per page.
- W3197758081 abstract "Visual defect detection is a key technology in modern industrial manufacturing systems. There are many possible appearances of product defects, including distortions in color, shape, contamination, missing or superfluous parts. For the detection of those, besides traditional image processing techniques, convolutional neural networks based methods have also appeared to avoid the usage of hand-crafted features and to build more efficient detection mechanisms. In our article we deal with autoencoder convolutional networks (AEs) which do not require examples of defects for training. Unfortunately, the manual and/or trial-and-error design of AEs is still required to achieve good performance, since there are many unknown parameters of AEs which can greatly influence the detection abilities. For our study we have chosen a well performing AE known as structural similarity AE (SSIM-AE), where the loss function and the comparison of the output with the input is implemented via the SSIM instead of the often used L1 or L2 norms. Investigating the performance of SSIM-AE on different data-sets, we found that its performance can be improved with modified convolutional structures without modifying the size of latent space. We also show that finding a model with low reconstruction error during training does not mean good detection abilities and denoising AEs can increase efficiency." @default.
- W3197758081 created "2021-09-13" @default.
- W3197758081 creator A5017910912 @default.
- W3197758081 creator A5044206520 @default.
- W3197758081 date "2021-01-01" @default.
- W3197758081 modified "2023-09-23" @default.
- W3197758081 title "About the Application of Autoencoders For Visual Defect Detection" @default.
- W3197758081 cites W2918626637 @default.
- W3197758081 cites W3118600296 @default.
- W3197758081 doi "https://doi.org/10.24132/csrn.2021.3101.20" @default.
- W3197758081 hasPublicationYear "2021" @default.
- W3197758081 type Work @default.
- W3197758081 sameAs 3197758081 @default.
- W3197758081 citedByCount "0" @default.
- W3197758081 crossrefType "proceedings-article" @default.
- W3197758081 hasAuthorship W3197758081A5017910912 @default.
- W3197758081 hasAuthorship W3197758081A5044206520 @default.
- W3197758081 hasBestOaLocation W31977580811 @default.
- W3197758081 hasConcept C101738243 @default.
- W3197758081 hasConcept C103278499 @default.
- W3197758081 hasConcept C108583219 @default.
- W3197758081 hasConcept C115961682 @default.
- W3197758081 hasConcept C119857082 @default.
- W3197758081 hasConcept C127413603 @default.
- W3197758081 hasConcept C153180895 @default.
- W3197758081 hasConcept C154945302 @default.
- W3197758081 hasConcept C21547014 @default.
- W3197758081 hasConcept C2778915421 @default.
- W3197758081 hasConcept C41008148 @default.
- W3197758081 hasConcept C81363708 @default.
- W3197758081 hasConceptScore W3197758081C101738243 @default.
- W3197758081 hasConceptScore W3197758081C103278499 @default.
- W3197758081 hasConceptScore W3197758081C108583219 @default.
- W3197758081 hasConceptScore W3197758081C115961682 @default.
- W3197758081 hasConceptScore W3197758081C119857082 @default.
- W3197758081 hasConceptScore W3197758081C127413603 @default.
- W3197758081 hasConceptScore W3197758081C153180895 @default.
- W3197758081 hasConceptScore W3197758081C154945302 @default.
- W3197758081 hasConceptScore W3197758081C21547014 @default.
- W3197758081 hasConceptScore W3197758081C2778915421 @default.
- W3197758081 hasConceptScore W3197758081C41008148 @default.
- W3197758081 hasConceptScore W3197758081C81363708 @default.
- W3197758081 hasLocation W31977580811 @default.
- W3197758081 hasLocation W31977580812 @default.
- W3197758081 hasOpenAccess W3197758081 @default.
- W3197758081 hasPrimaryLocation W31977580811 @default.
- W3197758081 hasRelatedWork W12007332 @default.
- W3197758081 hasRelatedWork W13382361 @default.
- W3197758081 hasRelatedWork W13618705 @default.
- W3197758081 hasRelatedWork W13953638 @default.
- W3197758081 hasRelatedWork W14516383 @default.
- W3197758081 hasRelatedWork W5460419 @default.
- W3197758081 hasRelatedWork W9190101 @default.
- W3197758081 hasRelatedWork W9563565 @default.
- W3197758081 hasRelatedWork W9750110 @default.
- W3197758081 hasRelatedWork W11830398 @default.
- W3197758081 isParatext "false" @default.
- W3197758081 isRetracted "false" @default.
- W3197758081 magId "3197758081" @default.
- W3197758081 workType "article" @default.