Don't skip the skips: autoencoder skip connections improve latent representation discrepancy for anomaly detection

Collin, Anne-Sophie;De Bodt, Cyril;Mulders, Dounia;De Vleeschouwer, Christophe
(2023) 31st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning — Location: Bruges, Belgium (4.October.2023)

Files

Collin2023_ES2023-139.pdf
  • Open Access
  • Adobe PDF
  • 2.93 MB
Download

Details

Authors
Abstract
Reconstruction-based anomaly detection typically relies on the reconstruction of a defect-free output from an input image. Such reconstruction can be obtained by training an autoencoder to reconstruct clean images from inputs corrupted with a synthetic defect. Previous works have shown that adopting an autoencoder with skip connections improves reconstruction sharpness. However, it remains unclear how skip connections affect the latent representations learned during training. Here, we compare internal representations of autoencoders with and without skip connections. Experiments over the MVTec AD dataset reveal that skip connections enable the autoencoder latent representations to intrinsically discriminate between clean and defective images.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Collin, A.-S., De Bodt, C., Mulders, D., & De Vleeschouwer, C. (2023). Don’t skip the skips: autoencoder skip connections improve latent representation discrepancy for anomaly detection. ESANN proceedings, 1(1), 1-6. https://doi.org/10.14428/esann/2023.ES2023-139 (Original work published 2023)