Transfer learning for Bag-of-Visual words approach to NBI endoscopic image classification

Annu Int Conf IEEE Eng Med Biol Soc. 2015 Aug:2015:785-8. doi: 10.1109/EMBC.2015.7318479.

Abstract

We address a problem of endoscopic image classification taken by different (e.g., old and new) endoscopies. Our proposed method formulates the problem as a constraint optimization that estimates a linear transformation between feature vectors (or Bag-of-Visual words histograms) in a framework of transfer learning. Experimental results show that the proposed method works much better than the case without feature transformation.

MeSH terms

  • Endoscopy*
  • Image Interpretation, Computer-Assisted