Multi-scale hybrid vision transformer and sinkhorn tokenizer for sewer defect classification

HIGHLIGHTS

• who: Joakim Bruslund Haurum from the Link to from University Visual Analysis and Perception Laboratory, University, Denmark have published the research: Multi-scale hybrid vision transformer and Sinkhorn tokenizer for sewer defect classification, in the Journal: (JOURNAL)
• what: Furtheru00ad more, the authors demonstrate that the Sinkhorn tokenizer successfully cluster the CNN features, which are expected to have a high amount of redunu00ad dant information due to successively applying overlapping convoluu00ad tional filters and pooling layers. The authors focus on the challenging task of multi-label sewer defect classification, which has been shown by Haurum and . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or E-mail:

  If you don't have an account, you can create one here.